Research article

Solvability and algorithm for Sylvester-type quaternion matrix equations with potential applications

  • Received: 07 January 2024 Revised: 08 May 2024 Accepted: 21 May 2024 Published: 19 June 2024
  • MSC : 15A09, 15A24, 15A29, 65F05

  • This article explores Sylvester quaternion matrix equations and potential applications, which are important in fields such as control theory, graphics, sensitivity analysis, and three-dimensional rotations. Recognizing that the determination of solutions and computational methods for these equations is evolving, our study contributes to the area by establishing solvability conditions and providing explicit solution formulations using generalized inverses. We also introduce an algorithm that utilizes representations of quaternion Moore-Penrose inverses to improve computational efficiency. This algorithm is validated with a numerical example, demonstrating its practical utility. Additionally, our findings offer a generalized framework in which various existing results in the area can be viewed as specific instances, showing the breadth and applicability of our approach. Acknowledging the challenges in handling large systems, we propose future research focused on further improving algorithmic efficiency and expanding the applications to diverse algebraic structures. Overall, our research establishes the theoretical foundations necessary for solving Sylvester-type quaternion matrix equations and introduces a novel algorithmic solution to address their computational challenges, enhancing both the theoretical understanding and practical implementation of these complex equations.

    Citation: Abdur Rehman, Ivan Kyrchei, Muhammad Zia Ur Rahman, Víctor Leiva, Cecilia Castro. Solvability and algorithm for Sylvester-type quaternion matrix equations with potential applications[J]. AIMS Mathematics, 2024, 9(8): 19967-19996. doi: 10.3934/math.2024974

    Related Papers:

    [1] T. K. Buvaneshwari, D. Anuradha . Solving bi-objective bi-item solid transportation problem with fuzzy stochastic constraints involving normal distribution. AIMS Mathematics, 2023, 8(9): 21700-21731. doi: 10.3934/math.20231107
    [2] Shima Soleimani Manesh, Mansour Saraj, Mahmood Alizadeh, Maryam Momeni . On robust weakly ε-efficient solutions for multi-objective fractional programming problems under data uncertainty. AIMS Mathematics, 2022, 7(2): 2331-2347. doi: 10.3934/math.2022132
    [3] Md. Musa Miah, Ali AlArjani, Abdur Rashid, Aminur Rahman Khan, Md. Sharif Uddin, El-Awady Attia . Multi-objective optimization to the transportation problem considering non-linear fuzzy membership functions. AIMS Mathematics, 2023, 8(5): 10397-10419. doi: 10.3934/math.2023527
    [4] Habibe Sadeghi, Fatemeh Moslemi . A multiple objective programming approach to linear bilevel multi-follower programming. AIMS Mathematics, 2019, 4(3): 763-778. doi: 10.3934/math.2019.3.763
    [5] N. Pazhaniraja, Shakila Basheer, Kalaipriyan Thirugnanasambandam, Rajakumar Ramalingam, Mamoon Rashid, J. Kalaivani . Multi-objective Boolean grey wolf optimization based decomposition algorithm for high-frequency and high-utility itemset mining. AIMS Mathematics, 2023, 8(8): 18111-18140. doi: 10.3934/math.2023920
    [6] Xin-Hui Shao, Wan-Chen Zhao . Relaxed modified Newton-based iteration method for generalized absolute value equations. AIMS Mathematics, 2023, 8(2): 4714-4725. doi: 10.3934/math.2023233
    [7] Muhammad Akram, Syed Muhammad Umer Shah, Mohammed M. Ali Al-Shamiri, S. A. Edalatpanah . Extended DEA method for solving multi-objective transportation problem with Fermatean fuzzy sets. AIMS Mathematics, 2023, 8(1): 924-961. doi: 10.3934/math.2023045
    [8] Sema Akin Bas, Beyza Ahlatcioglu Ozkok . An iterative approach for the solution of fully fuzzy linear fractional programming problems via fuzzy multi-objective optimization. AIMS Mathematics, 2024, 9(6): 15361-15384. doi: 10.3934/math.2024746
    [9] Guangjian Li, Guangjun He, Mingfa Zheng, Aoyu Zheng . Uncertain multi-objective dynamic weapon-target allocation problem based on uncertainty theory. AIMS Mathematics, 2023, 8(3): 5639-5669. doi: 10.3934/math.2023284
    [10] Ya Qin, Rizk M. Rizk-Allah, Harish Garg, Aboul Ella Hassanien, Václav Snášel . Intuitionistic fuzzy-based TOPSIS method for multi-criterion optimization problem: a novel compromise methodology. AIMS Mathematics, 2023, 8(7): 16825-16845. doi: 10.3934/math.2023860
  • This article explores Sylvester quaternion matrix equations and potential applications, which are important in fields such as control theory, graphics, sensitivity analysis, and three-dimensional rotations. Recognizing that the determination of solutions and computational methods for these equations is evolving, our study contributes to the area by establishing solvability conditions and providing explicit solution formulations using generalized inverses. We also introduce an algorithm that utilizes representations of quaternion Moore-Penrose inverses to improve computational efficiency. This algorithm is validated with a numerical example, demonstrating its practical utility. Additionally, our findings offer a generalized framework in which various existing results in the area can be viewed as specific instances, showing the breadth and applicability of our approach. Acknowledging the challenges in handling large systems, we propose future research focused on further improving algorithmic efficiency and expanding the applications to diverse algebraic structures. Overall, our research establishes the theoretical foundations necessary for solving Sylvester-type quaternion matrix equations and introduces a novel algorithmic solution to address their computational challenges, enhancing both the theoretical understanding and practical implementation of these complex equations.



    As one of the extensions of artificial neural networks (ANNs), the cellular neural network (CNN) was initially proposed by Chua and Yang in [1,2]. The connections of neurons in CNNs are localized, which means that neurons in CNNs only connect to other neurons in a certain range. This local connection method not only reduces the connection between neurons, but also improves the efficiency of parallel computing. Based on this, CNNs can be used in many fields, such as image encryption [3], parallel computing [2], and so on [4,5].

    However, for some problems which have more complexity and vagueness, the models established by using general neural networks are often not accurate enough. Hence, the Takagi-Sugeno fuzzy model (TSFM) [6] as an effective tool to solve this problem has received widespread attention in recent years. The TSFM is a nonlinear system description approach that employs linear weighted inequalities and differential equations, and TSFMs can simulate any smooth dynamical system in any precision which are defined on a compact set by connecting several local linear models with fuzzy membership functions piecewise smoothly. Based on these properties, very recently, TSFMs have been used to describe CNNs, e.g., T-S fuzzy CNN (TSFCNN) is presented [7,8,9,10]. TSFCNNs combine the advantages of fuzzy theory and features of neural networks which improve the ability of general CNNs to handle complex problems by using the IF-THEN rules. Therefore, TSFCNNs are widely used in more fields than CNNs.

    For a designed neural network model, stability is one of the indicators that determines whether it has good performance. However, throughout the real simulation procedure, due to the inevitable existence of disturbances, such as times delays and stochastic disturbances, the stability of TSFCNNs may not be achieved. Time delays are usually caused by the limited signal transmission speed of electronic devices, and different ways of operation of electrical devices might result in varying delays., such as time-varying delays [11], state-dependent delays [12] or distributed delays [13] etc. In addition, stochastic disturbance is a type of complicated and irregular disruption, and it is distinct from traditional processes. The intensities of both of the above disturbances can to some extent make the designed TSFCNN model unable to achieve the expected performance. Hence, the stability analysis of TSFCNNs have received more and more attention in recent decades [14,15,16,17,18]. In [14], Hou et al. analyzes the global exponential stability of delayed TSFCNNs (DTSFCNNs) by constructing Lyapunov-Krasovskii functionals as well as using the linear matrix inequality (LMI) method. In [15,18], Balasubramaniam et al. derived some criteria based on the LMI method to ensure the stability of stochastic TSFCNNs (STSFCNNs) and DTSFCNNs, respectively. In [16], Yang and Sheng discuss the robust stability of uncertain STSFCNNs in detail. Long and Xu further considered the impacts of impulses based on delayed STSFCNNs (DSTSFCNNs) in [17], and explore its exponential p-stability by using the property of M-matrix.

    It is worth mentioning that the above literature is all about the stability analysis of DTSFCNNs or STSFCNNs rather than robustness of stability. Robustness is the ability of a system to maintain its dynamic properties under parameter changes. Once the parameters change beyond a certain fixed range, the original dynamic properties of the system will change. In the actual modeling process, we always want to design a model with good robustness. However, few researchers have described the range of parameter variations that the systems can withstand. For the problem of the robustness of stability (RoS), the classical methods will no longer be applicable. Hence, Shen et al. first proposed an inequality method based on the Gronwall-Bellman lemma to explore the robustness of stability in [19]. Based on this method, many results which worked on the problem of RoS of neural networks are presented, for example, Si et al. investigated the RoS of the dynamical systems with deviating arguments based on Gronwall-Bellman lemma in [20,21]. Moreover, Fang et al. further explore the RoS of a class of CNNs based on bidirectional associative memory in [22].

    However, it should be pointed out that the above studies did not take into account the fuzziness in the modeling process, and did not explore the robustness of neural networks containing T-S fuzzy rules. Therefore, in order to fill this gap, this paper mainly analyzes the RoS of perturbed TSFCNNs. Key works and contributions of this paper are listed below.

    ● Time delay and random disturbance are two kinds of system disturbances mainly considered in this paper, both of which are unavoidable in practical modeling. Sufficient conditions are given in [9,12,14] for the stability of systems containing perturbations, but few have explored the upper bound on the strength of these perturbations. When the values of the disturbances change, the stability of the system will also change, so it is meaningful to discuss the maximum value of the disturbance that the system can withstand. The innovation of this paper lies in the following aspects: 1. This paper establishes a recurrent neural network with T-S fuzzy rules, and gives the robustness analysis of the model based on the classical Gronwall inequality. 2. The constraint relationship between time delay and random disturbance is established, which means that the relationship between time delay and random disturbance is dynamic. Besides, the results obtained in this paper extend the results in [19], which means that when there is only one fuzzy rule in DTSFCNNs or DSTSFCNNs, the results in [19] can be obtained.

    ● In comparison to [19,20,21], the subsystems considered in this paper are interconnected by fuzzy rules, and the parameters of each subsystem are not the same. With the increase of fuzzy rules, the number of subsystems increases, which greatly enhances the fuzziness of the networks considered in the literature in dealing with practical problems, and the existence of fuzzy rules also increases the difficulty of system analysis to a certain extent.

    ● The methods used in this paper mainly include inequality techniques, random analysis and algebraic knowledge, so the results obtained in this paper are easy to verify. The results derived in this paper are helpful for designing TSFCNNs with better performance.

    Finally, the organization of this paper is listed below. We introduce the models we considered and assumptions we needed in Section 2. The problem of the RoS of DTSFCNNs is explored in Section 3. In Section 4, we discuss the RoS of DSTSFCNNs. And, in Section 5, several examples are provided to verify the results in this paper.

    Notations: Z+={a|a=1,2,,w}. R=(,+), R+=(0,+), Rw and Rw×v are the sets of w-dimensional vectors, and w×v dimensional matrices, respectively. |||| denotes the Euclidean norm of any vectors or matrices. Define (ζ,F,{Ft}t0,P) as the complete filtered probability space containing all P-null sets, where right continuous filtration Ft0 meets the usual conditions. Scalar Brownian movement ς(t) is defined in (ζ,F,{Ft}t0,P). L2F0([μ,0];Rw) is the set which contains all F0 measurable C([μ,0];Rw) valued stochastic variables ={(ι):μι0} and supμι0E||(ι)||2. E is the expectation operator. μ=max{μ1,μ2,,μn}R+, where μj is a constant delay. AT represents the transposition of matrix or vector A.

    We consider the delayed CNNs which is in the following form:

    {˙zh(t)=chzh(t)+nu=1ahuyu(zu(t))+nu=1bhuyu(zu(tμu))+Jh,zh(t)=ϕh(t),t[μ,0], (2.1)

    where ch represents the rates of hth neuron resetting its potential to the isolated resting state, and ch>0. ahu and bhu are the strengths of uth neuron on the hth neuron at time t and tμu, where μu represents the transmission delay, which is a positive constant. Jh represents the external bias which impacts the hth neuron.

    Remark 2.1. Time delays are often present in real-world control systems, and there may be delays in states, control inputs or measurements. Delay will introduce uncertainty and interference, which will degrade the robustness of the controller. Therefore, the effect of time delay must be considered when designing the control system.

    Let C=diag{c1,c2,,cn}, A=(ahu)n×n, B=(bhu)n×n, y()={y1(),y2(),,yn()}T, and ν={μ1,μ2,,μn}T. Then, CNN (2.1) has the following representation:

    {˙z(t)=Cz(t)+Ay(z(t))+By(z(tν))+J,z(t)=ϕ(t),t[μ,0]. (2.2)

    Assume the equilibrium point is z={z1,z2,,zn}, and ρi(t)=zi(t)zi. Then, CNN (2.2) is equivalent to the following model

    {˙ρ(t)=Cρ(t)+Ag(ρ(t))+Bg(ρ(tν)),ρ(t)=ψ(t),t[μ,0], (2.3)

    where g(ρ(t))=y(ρ(t)+z)y(z), g(ρ(tν))=y(ρ(tν)+z)y(z) and ψ(t)=ϕ(t)z.

    Remark 2.2. The circuit implementations of (2.1)–(2.3) can be found in [1,23].

    The following condition is one we assume the function g() needs to satisfy:

    Assumption 2.1. There is an L>0 such that

    ||g(u)g(v)||L||uv|| (2.4)

    holds, where u and v are two states of the CNN (2.3).

    Remark 2.3. Since g(ρ(t))=y(ρ(t)+z)y(z), hence, g(0)=0. Therefore, from Assumption 2.1, ||g(u)||L||u|| holds, i.e., the linear growth condition for activation function g() is also satisfied.

    Remark 2.4. There are many functions that satisfy the Assumption 2.1, such as g(x)=tanh(x), g(x)=(|x+1||x1|)/2 and so on. The selection of activation functions is closely related to the stability of CNNs. For example, consider the following neural network model

    ˙x(t)=2x(t)+1.1f(x(t))+f(x(tτ)), (2.5)

    where τ=3. We take the activation functions as f(x)=tanh(x) and f(x)=(|x+1||x1|)/2, respectively. It can be seen from Figure 1 that when the model has different activation functions, its convergence rate to reach the stable state is affected.

    Figure 1.  States of system (2.5) with different activation functions.

    By applying fuzzy IF-THEN rules to the CNN (2.3), the following DTSFCNN model can be obtained:

    PlantRulek:IFϖ1(t)isυk1andϖ2(t)isυk2andandϖm(t)isυkmTHEN{˙ρ(t)=Ckρ(t)+Akg(ρ(t))+Bkg(ρ(tν)),ρ(t)=ψ(t),t[μ,0],k=1,2,,l, (2.6)

    where ϖj(t) and υkj are the premise variable and the fuzzy set respectively. l represents the amount of fuzzy IF-THEN rules. Ck, Ak, Bk are constant matrices.

    Remark 2.5. By combining T-S fuzzy logic with nonlinear differential equations, the described system simulates a smooth system with arbitrary precision by using fuzzy rules to connect several nonlinear models with fuzzy membership functions piecewise and smoothly on compact sets.

    Remark 2.6. Compared with the existing literature [19,20,21], the system with T-S fuzzy logic considered in this paper is composed of multiple sets of nonlinear systems connected by T-S fuzzy logic, and the parameters of each nonlinear system are not fixed, which can more accurately describe the changes of system parameters brought by fuzziness during the operation of the actual dynamic system.

    The DTSFCNN (2.6) after being defuzzied is shown as follows:

    ˙ρ(t)=lk=1Ωk(ϖ(t)){Ckρ(t)+Akg(ρ(t))+Bkg(ρ(tν))}, (2.7)

    where ϖ(t)={ϖj(t)}1×m and

    Ωk(ϖ(t))=γk(ϖ(t))lk=1γk(ϖ(t)),γk(ϖ(t))=mj=1υkj(ϖj(t)). (2.8)

    Moreover, Ωk(ϖ(t)) represents the averaged weight of the rules. υkj(ϖj(t)) is the grade of the membership function of ϖj(t) in υkj. From fuzzy set theory, we can get

    γk(ϖ(t))0,lk=1γk(ϖ(t))>0, (2.9)

    and

    Ωk(ϖ(t))0,lk=1Ωk(ϖ(t))=1, (2.10)

    where k=1,2,,l.

    When μ=0, DTSFCNN (2.6) degenerates into the following TSFCNN.

    PlantRulek:IFϖ1(t)isυk1andϖ2(t)isυk2andandϖm(t)isυkmTHEN{˙κ(t)=Ckκ(t)+Akg(κ(t))+Bkg(κ(t)),κ(t0)=ψ(t0),k=1,2,,l. (2.11)

    Then, we can get the following defuzzied output of the above TSFCNN.

    ˙κ(t)=lk=1Ωk(ϖ(t)){Ckκ(t)+Akg(κ(t))+Bkg(κ(t))}. (2.12)

    Next, we give the definition of global exponential stability of system (2.7).

    Definition 2.1. A DTSFCNN (2.7) is globally exponentially stable (GES) if

    ||ρ(t)||αexp(β(tt0))sups[t0μ,t0]||ρ(s)|| (2.13)

    holds, where α, β are positive constants.

    Remark 2.7. When μ=0, system (2.7) will degenerate into the ideal delay-free system (2.12), and Definition 2.1 can be rewritten as ||ρ(t)||αexp(β(tt0))||ρ(t0)||. In this paper, unless otherwise stated, TSFCNNs (2.12) are GES, i.e., ||κ(t)||αexp(β(tt0))||κ(t0)|| holds t>t0.

    Another assumption is needed to obtain our major results.

    Assumption 2.2. There are positive constants α, β and ð such that

    2m2exp(2(m1+m2)ð)+αexp(βð)<1 (2.14)

    holds, where m1=lk=1||Ck||+Llk=1 and m2=lk=1||Bk||.

    Next, let us analyze definition of global exponential stability of DTSFCNNs (2.6) first.

    Theorem 3.1. Let Assumptions 2.1 and 2.2 hold, then a DTSFCNN (2.6) is GES when μmin{ð/2,ˆμ}, where ˆμ satisfies the following equation.

    m2{ˆμ[α/β(m1+m2)+m2]+2}exp{2ð[m1+m2+m2ˆμ(m1+m2)]}+αexp(β(ðˆμ))=1, (3.1)

    where ð>lnα/β, m1=lk=1||Ck||+lk=1L||Ak||, and m2=lk=1L||Bk||.

    Proof. From (2.6) and (2.11),

    κ(t)ρ(t)=tt0lk=1Ωk(ϖ(ς)){Ck(κ(ς)ρ(ς))+Ak(g(κ(ς))g(ρ(ς)))+Bk(g(κ(ς))g(ρ(ςν)))}dς. (3.2)

    According to (2.10), we can get Ωk(ϖ(t))1. Therefore,

    ||κ(t)ρ(t)||tt0{(lk=1||Ωk(ϖ(ς))Ck||+Llk=1||Ωk(ϖ(ς))Ak||)×||κ(ς)ρ(ς)||+Llk=1||Ωk(ϖ(ς))Bk||||κ(ς)ρ(ςν)||}dςtt0m1||κ(ς)ρ(ς)||+m2||κ(ς)ρ(ςν)||dςtt0(m1+m2)||κ(ς)ρ(ς)||+m2||ρ(ς)ρ(ςν)||dς, (3.3)

    where m1=lk=1||Ck||+lk=1L||Ak||, and m2=lk=1L||Bk||.

    Since

    ρ(ς)ρ(ςν)ςςμ˙ρ(s)ds, (3.4)

    therefore,

    ||ρ(ς)ρ(ςν)||ςςμ||˙ρ(s)||dsςςμm1||ρ(s)||+m2||ρ(sν)||ds. (3.5)

    Thus, when t>t0+μ,

    tt0+μ||ρ(ς)ρ(ςν)||dςtt0+μςςμm1||ρ(s)||+m2||ρ(sν)||dsdςtt0min{s+μ,t}max{t0+μ,s}m1||ρ(s)||+m2||ρ(sν)||dςdsμtt0m1||ρ(s)||+m2||ρ(sν)||dsμtt0m1||ρ(s)κ(s)||ds+μtt0m1||κ(s)||ds+μtt0m2||ρ(sν)||dsm1μtt0||ρ(s)κ(s)||ds+m1μαβsup1||κ(s)||+m2μsup1||ρ(s)||+m2μtt0||ρ(s)||dsμ(m1+m2)tt0||ρ(s)κ(s)||ds+μ[α/β(m1+m2)+m2]sup1||ρ(s)||, (3.6)

    where sup1=sups[t0μ,t0].

    On the other hand,

    t0+μt0||ρ(ς)ρ(ςν)||dςsup2||ρ(s)||+sup1||ρ(s)||, (3.7)

    where sup2=sups[t0,t0+μ].

    Thus, we have

    tt0||ρ(ς)ρ(ςν)||dςμ(m1+m2)tt0||ρ(s)κ(s)||ds+μ[α/β(m1+m2)+m2]sup1||ρ(s)||+sup2||ρ(s)||+sup1||ρ(s)||μ(m1+m2)tt0||ρ(s)κ(s)||ds+{μ[α/β(m1+m2)+m2]+2}sup3||ρ(s)||, (3.8)

    where sup3=sups[t0μtt0+μ].

    Then, from (3.3),

    ||κ(t)ρ(t)||{m1+m2+m2μ(m1+m2)}tt0||ρ(s)κ(s)||ds+m2{μ[α/β(m1+m2)+m2]+2}sup3||ρ(s)||=:m3tt0||ρ(s)κ(s)||ds+m4sup3||ρ(s)||. (3.9)

    By applying the Gronwall-Bellman lemma, when tt0+2ð,

    ||κ(t)ρ(t)||m4exp(2ðm3)sup3||ρ(s)||. (3.10)

    Therefore, for t0+μtt0+2ð,

    ||ρ(t)||||κ(t)ρ(t)||+||κ(t)||m4exp(2ðm3)sup3||ρ(s)||+αexp(β(tt0))sup1||ρ(s)||. (3.11)

    Notice that μð/2, and thus, for t0μ+ðt0μ+2ð, we have

    ||ρ(t)||{m4exp(2ðm3)+αexp(β(ðμ))}sup4||ρ(s)||. (3.12)

    where sup4=sups[t0μtt0μ+ð].

    Let F(μ)=m4exp(2ðm3)+αexp(β(ðμ)). Then, we can easily get that F is strictly increasing with respect to μ, and F(0)<1. Hence, there must be a ˆμ such that F(ˆμ)=1 holds, i.e., when μmin{ð/2,ˆμ}, F(μ)<1 holds.

    Take Γ=lnF/ð. Then, from (3.12), we can obtain

    sup5||ρ(s)||exp(Γð)sup4||ρ(s)||, (3.13)

    where sup5=sups[t0μ+ð,t0μ+2ð].

    Then, there is a ξZ+ such that

    λ=sups[t0+ξðμ,t0+(ξ+1)ðμ]||ρ(s)||exp(Γð)sups[t0+(ξ1)ðμ,t0+ξðμ]||ρ(s)||exp(Γξð)sup5||ρ(s)||. (3.14)

    Clearly, for t>t0+ðμ,

    ||ρ(s)||exp(Γð)exp(Γ(tt0))λ. (3.15)

    We can verify this is also true for t0tt0+ðμ. Thus, the DTSFCNN (2.6) is GES when μ<min{ð/2,ˆμ}.

    Remark 3.1. Compared with the existing results [16,19,20,21], this paper combines T-S fuzzy rules with nonlinear differential equations, which makes the system similar to a switching system, where each fuzzy rule corresponds to a definite subsystem. On this basis, the addition of fuzzy rules allows nonlinear systems to be fitted using combinations of linear systems.

    This section will mainly consider the RoS of TSFCNN under the combined effects of time delays and stochastic disturbances.

    We consider the following DSTSFCNNs under the same fuzzy rules of DTSFCNNs (2.6).

    PlantRulek:IFϖ1(t)isυk1andϖ2(t)isυk2andandϖm(t)isυkmTHEN{d(t)=[Ck(t)+Akg((t))+Bkg((tν))]dt+Hk(t)dς(t),(t)=ψ(t),t[μ,0],k=1,2,,l, (4.1)

    where represents the intensity of stochastic disturbances. Hk is a constant matrix.

    Remark 4.1. Since the errors are generated by rounding when doing numerical calculations, although these errors are determined, they are actually impossible to calculate. The way to overcome this phenomenon is to model these errors as small random perturbations of the system, for example, large-scale deterministic systems used to model climate evolution may exhibit "noise-like" characteristics in some subsystems on short time scales. By using noise to model these subsystems, calculations can be greatly accelerated. Hence, under normal circumstances, random factors must be considered in the system to be more realistic.

    Similarly, the defuzzied model is as follows.

    d(t)=lk=1Ωk(ϖ(t)){[Ck(t)+Akg((t))+Bkg((tν))]dt+Hk(t)dς(t)}. (4.2)

    Next, we introduce the definition exponential stability of DSTSFCNN (4.1) in mean square.

    Definition 4.1. DSTSFCNN (4.1) is mean square exponentially stable (MSES) if there exist ω>0, ϱ>0, such that

    E||(t)||2ω2sups[t0μ,t0]E||(s)||2exp(2ϱ(tt0)) (4.3)

    holds.

    Theorem 4.1. Let Assumption 2.1 hold, DSTSFCNN (4.1) is MSES if ||<ˆ/2 and μ<min{ð/2,ˆμ}, where ˆ and ˆμ satisfies two equations listed below, respectively.

    12ˆ2m25exp{2ð[6ð(m21+2m22)+6ˆ2m25]}+2α2exp(2βð)=1, (4.4)
    2{12ðm22[ˆμ[3ˆμm21+3m25ˆ2/2]α2/β+3ˆμ2m22+4ˆμ]+3ˆ2m25}×exp{2ð[6ð(m21+2m22)+3ˆ2m25+12ðm22(6ˆμ(ˆμm21+m25ˆ2/2))]}+2α2exp(2β(ðˆμ))=1, (4.5)

    where m1 and m2 are defined in Theorem 3.1, and m5=lk=1||Hk||.

    Proof. From (4.2) and (2.7), we can obtain

    (t)κ(t)=tt0lk=1Ωk(ϖ(s)){Ck((s)κ(s))+Ak(g((s))g(κ(s)))+Bk(g((sν))g(κ(s)))}ds+tt0lk=1Ωk(ϖ(t))[Hk(s)]dς(s). (4.6)

    Thus,

    ||(t)κ(t)||tt0m1||(s)κ(s)||+m2||(sν)κ(s)||dt+tt0m5||(s)||dς(s), (4.7)

    where m5=lk=1||Hk||.

    By utilizing the Itˆo formula, when tt0+2ð,

    E||(t)κ(t)||26ðm21tt0E||(s)κ(s)||2ds+6ðm22tt0E||(sν)κ(s)||2ds+32m25tt0E||(s)||2ds6ð(m21+2m22)tt0E||(s)κ(s)||2ds+12ðm22tt0E||(sν)(s)||2ds+62m25tt0E||(s)κ(s)||2+E||κ(s)||2ds[6ð(m21+2m22)+62m25]tt0E||(s)κ(s)||2ds+12ðm22tt0E||(sν)(s)||2ds+62m25tt0E||κ(s)||2ds. (4.8)

    When t>t0+μ,

    tt0+μE||(sν)(s)||2dstt0+μssμ{[3μm21+32m25]E||(s)||2+3μm22E||(sν)||2}dϑdsmin{ϑ+μ,t}max{ϑ,t0+μ}tt0{[3μm21+32m25]E||(s)||2+3μm22E||(sν)||2}dsdϑμ[3μm21+32m25]tt0E||(s)||2ds+3μ2m22tt0E||(sν)||2ds. (4.9)

    Let m6=6μ[μm21+2m25], and m7=3μ2m22. Thus, we can obtain

    tt0+μE||(sν)(s)||2dsm6tt0E||(s)κ(s)||2ds+m6tt0E||κ(s)||2ds+m7t0t0μE||(s)||2dsm6tt0E||(s)κ(s)||2ds+(m6α22β+m7)sup1E||(s)||2, (4.10)

    where sup1=sups[t0μ,t0].

    In addition,

    t0+μt0E||(sν)(s)||2ds2μsup1E||(s)||2+2μsup2E||(s)||2, (4.11)

    where sup2=sups[t0,t0+μ].

    Hence,

    E||(t)κ(t)||2[6ð(m21+2m22)+62m25]tt0E||(s)κ(s)||2ds+12ðm22tt0E||(sν)(s)||2ds+62m25tt0E||κ(s)||2ds[6ð(m21+2m22)+62m25]tt0E||(s)κ(s)||2ds+12ðm22[m6tt0E||(s)κ(s)||2ds+(m6α22β+m7)sup1E||(s)||2+4μsup3E||(s)||2]+62m25tt0E||κ(s)||2ds=:m8tt0E||(s)κ(s)||2ds+m9sup3E||(s)||2, (4.12)

    where

    m8=6ð(m21+2m22)+62m25+12ðm22m6,m9=12ðm22(m6α2/2β+m7+4μ)+32m25α2/β,sup3=sups[t0μ,t0+μ].

    Therefore, when t0+μtt0+2ð, using the Gronwall-Bellman lemma, we can get

    E||(t)κ(t)||2m9exp(2m8ð)sup3E||(s)||2. (4.13)

    Then, for t0μ+ðtt0μ+2ð

    E||(t)||22E||(t)κ(t)||2+2E||κ(t)||22[m9exp(2m8ð)+α2exp(2β(ðμ))]sup4E||(s)||2, (4.14)

    where sup4=sups[t0μ,t0μ+ð].

    Select (μ,)=m9exp(2m8ð)+α2exp(2β(ðμ)). Then, (0,) is strictly increasing with respect to . Hence, there exists an ˆ>0 such that (0,ˆ)=1. Besides, when =2, (μ,/2) is also strictly increasing for μ, and thus there exists a ˆμ such that (ˆμ,ˆ/2)=1. That is, (μ,)<1 when μmin{ð/2,ˆμ} and ˆ/2.

    Furthermore, the remaining part of the proof is exactly the same with Theorem 3.1, and hence it is omitted here. Hence, a DSTSFCNN (4.1) is MSES when μmin{ð/2,ˆμ}, and ˆ/2.

    Remark 4.2. Figure 2 shows the detailed analysis steps of Theorems 3.1 and 4.1. In addition, in the derivation of Theorem 4.1, due to the existence of random perturbations, the number of terms of the system increases, which also makes the inequality used in Theorem 4.1 and Theorem 3.1 have some differences, so Theorem 4.1 is not a simple superposition of Theorem 3.1. Besides, if there is only one fuzzy rule in (2.3) and (4.2), then (2.3) and (4.2) can be seen as certain neural networks, and we can obtain the Theorems 2 and 3 of [19] respectively.

    Figure 2.  The steps of the analysis of Theorems 1 and 2.

    Remark 4.3. Table 1 provides a comparison of the existing literature with this paper. Elements to be compared are time delays (T-D), stochastic disturbances (S-D), RoS, T-S type fuzzy model (T-S fuzzy), global asymptotic stability (GAS), GES and MSES.

    Table 1.  The differences between this paper and several existing literature.
    T-S fuzzy T-D S-D RoS GAS GES MSES
    Singh et al. [9] (2021) - - - - - -
    Li et al. [12] (2022) - - - - -
    Almarri et al. [13] (2022) - - - - - -
    Si et al. [20] (2021) - - -
    Wenxiang et al. [22] (2023) - - -
    Zhang et al. [24] (2023) - - -
    This paper -

     | Show Table
    DownLoad: CSV

    Example 1. We take the following DTSFCNN model into account:

    PlantRule1:IFϖ1(t)isυ11andϖ2(t)isυ12THEN{˙ρ(t)=C1ρ(t)+A1g(ρ(t))+B1g(ρ(tν)),ρ(t)=ψ(t),t[μ,0],PlantRule2:IFϖ1(t)isυ21andϖ2(t)isυ22THEN{˙ρ(t)=C2ρ(t)+A2g(ρ(t))+B2g(ρ(tν)),ρ(t)=ψ(t),t[μ,0], (5.1)

    where

    C1=[1001.2],C2=[1.1001],A1=[0.010.010.010.01],A2=[0.010.010.010.01],B1=[0.010.010.010.01],B2=[0.010.010.010.01].

    Taking the follow functions as the fuzzy membership functions:

    Ω1(ϖ(t))=γ1(ϖ(t))2k=1γk(ϖ(t)),Ω2(ϖ(t))=γ2(ϖ(t))2k=1γk(ϖ(t)),

    where 1=ρ1(t)+3,2=ρ2(t)+3, and

    γ1(ϖ(t))=exp[211.4426]exp[221.4426],γ2(ϖ(t))=exp[211.4426]exp[(26)21.4426].

    Let g()=tanh(), α=1, β=0.9, ð=0.2 and L=1. Then, we can obtain that

    0.0566exp(0.9450)+exp(0.18)=0.9808<1, (5.2)

    and from Theorem 3.1, we have

    (0.075ˆμ+0.0566)exp(0.4(2.3624+0.0668ˆμ))+exp(0.9(0.2ˆμ))=1. (5.3)

    Therefore, we can get that ˆμ=0.0201 by solving Eq (5.3). Thus, the DTSFCNN (5.1) is GES when μ<min{ð/2,ˆμ}.

    Figure 3 shows the relationship between the states of the DTSFCNN (5.1) and the membership functions Ωk(ϖ(t)). We can find that, as the state values decrease, the values of both membership functions tend to approach 0.5. In Figure 4, we choose μ=0.015<0.0201, and then the DTSFCNN is GES.

    Figure 3.  Values of membership functions Ωk(ϖ(t)).
    Figure 4.  The states of the DTSFCNN (5.1) with μ=0.015 for different initial values.

    Example 2. The DSTSFCNN model we consider is of the following form:

    PlantRule1:IFϖ1(t)isυ11andϖ2(t)isυ12THEN{d(t)=[C1(t)+A1g((t))+B1g((tν))]dt+H1(t)dς(t),(t)=ψ(t),t[μ,0],PlantRule2:IFϖ1(t)isυ21andϖ2(t)isυ22THEN{d(t)=[C2(t)+A2g((t))+B2g((tν))]dt+H2(t)dς(t),(t)=ψ(t),t[μ,0], (5.4)

    where membership function Ωk(ϖ(t)) is the same as that in Example 1.

    Accordingly, the following is the defuzzied DSTSFCNN (5.4).

    d(t)=lk=1Ωk(ϖ(t)){[Ck(t)+Akg((t))+Bkg((tν))]dt+Hk(t)dς(t)}, (5.5)

    where

    C1=[1001],C2=[1001],A1=[0.010.010.010.01],A2=[0.010.010.010.01],B1=[0.010.010.010.01],B2=[0.010.010.010.01],H1=[0.10.10.10.1],H2=[0.10.10.10.1].

    Take α=1, β=0.9, ð=0.4 and L=1. Then, we can get the following two equations from Theorem 4.1:

    1.92ˆ2exp(7.9032+0.7680ˆ2)+0.9735=1, (5.6)

    and

    {0.0112[1.1111ˆμ[12.3418ˆμ+0.24ˆ2]+0.0035ˆμ2+4ˆμ]+1.0667ˆ2}×exp{7.9032+0.3840ˆ2+0.0045(6ˆμ(4.1189ˆμ+0.08ˆ2))}+2exp(1.8(0.4ˆμ))=1. (5.7)

    Take ζ(μ,)=(μ,)1, ˆζ()=(0,)1 and ˉζ(μ)=(μ,ˆ/2)1. Then, the problem of solving Eqs (5.6) and (5.7) can be transformed into finding the zero point of functions ˆζ() and ˉζ(μ). In addition, Figure 5 shows the values of functions ζ(μ,), ˆζ() and ˉζ(μ), respectively.

    Figure 5.  Values of functions ζ(μ,), ˆζ() and ˉζ(μ).

    From Figure 5b, c and Eqs (5.6) and (5.7), we can easily obtain that ˆ=0.0030 and ˆμ=1.0777×104. This means that the system can remain globally exponentially stable when the perturbation intensity is less than the upper bounds derived in this paper.

    In Figure 6, we take =0.0001<0.003 and μ=0.0001<1.0777×104, and hence, the states of (5.5) with different initial values are MSES. Figure 7 depicts the states of (5.5) in the sense of mean square. When time delay and stochastic noise intensity are all bigger than the results we derive, that is, when =0.01 and μ=0.001, (5.5) exhibits unstable behavior and cannot maintain global exponential stability. This can easily be seen in Figure 8.

    Figure 6.  The states of (5.5) with =0.0001, μ=0.0001 for different initial values.
    Figure 7.  The values of (5.5) with =0.0001, μ=0.0001 in the sense of mean square.
    Figure 8.  The states of (5.5) with ˆμ=0.001, ˆ=0.01 for different initial values.

    Figures 9 and 10 show that when the strength of one of the disturbance factors is greater than the upper bound derived by Theorem 4.1, the system still cannot maintain global exponential stability. Therefore, it can be seen that the system can maintain exponential stability only when the strength of both disturbances is smaller than the upper bound derived in this paper. This also fully proves the validity of the results obtained in this paper.

    Figure 9.  The states of (5.5) with ˆμ=0.0001, ˆ=0.01 for different initial values.
    Figure 10.  The states of (5.5) with ˆμ=0.1, ˆ=0.0001 for different initial values.

    Remark 5.1. These examples show that if the requirements in Theorems 3.1 and 4.1 are met, the perturbed systems are globally exponentially stable. The transcendental equations satisfied by the maximum delay and maximum noise intensity are derived by inequality technology, and they can be easily solved by MATLAB, so the conditions in the theorem are easy to verify. In addition, Figure 11 shows the brief calculation process of Examples 1 and 2.

    Figure 11.  The calculation process of Examples 1 and 2.

    Remark 5.2. Since the inequality technique is mainly used in this paper, error is inevitable. Therefore, the upper bound of the perturbation obtained in this paper is only a sufficient condition for the system to maintain exponential stability. Furthermore, due to the monotonically increasing nature of the exponential function, this leads to the fact that Assumption 2.2 is not easily satisfied. Future research will focus on using the new inequalities to weaken the conservativeness of Assumption 2.2 and further derive more accurate upper bounds of perturbations.

    In short, we analyze the robustness of the stability of recurrent neural networks with T-S fuzzy rules. First, we decouple the proposed neural network models. Second, we discuss the maximum delay length and maximum noise intensity of DTSFCNNs and DSTSFCNNs by using the inequality technique, stochastic analysis and algebraic methods. Moreover, numerical examples show that the disturbed system can still maintain exponential stability when the intensity of the perturbations is lower than the results derived in this paper.

    Future research may concentrate on merging well-known methodologies such as the LMI method, Lyapunov theory, and so on to lessen the conservativeness of this work. Besides, on the basis of this paper, we can continue to consider the uncertainties in the modeling, such as deviating parameters and Markov switching. We can also continue to consider more complex neural network models, such as higher-order neural networks, neutral neural networks, etc.

    The authors declare they have not used Artificial Intelligence (AI) tools in the creation of this article.

    The authors declare that there are no conflicts of interest regarding the publication of this paper.



    [1] R. A. Horn, C. R. Johnson, Matrix analysis, Cambridge: Cambridge University Press, 2012. https://doi.org/10.1017/CBO9780511810817
    [2] K. Zhou, J. C. Doyle, K. Glover, Robust and optimal control, Upper Saddle River: Prentice Hall, 1996. Available from: https://dl.acm.org/doi/book/10.5555/225507
    [3] V. Simoncini, Computational methods for linear matrix equations, SIAM Rev., 58 (2016), 377–441. https://doi.org/10.1137/130912839 doi: 10.1137/130912839
    [4] V. L. Syrmos, F. L. Lewis, Coupled and constrained Sylvester equations in system design, Circuits Syst. Signal Process., 13 (1994), 663–694. https://doi.org/10.1007/BF02523122 doi: 10.1007/BF02523122
    [5] K. R. Gavin, S. P. Bhattacharyya, Robust and well-conditioned eigenstructure assignment via Sylvester's equation, Proc. Amer. Control Conf., 1982. https://doi.org/10.1002/oca.4660040302
    [6] M. Darouach, Solution to Sylvester equation associated to linear descriptor systems, Syst. Control. Lett., 55 (2006), 835–838. https://doi.org/10.1016/j.sysconle.2006.04.004 doi: 10.1016/j.sysconle.2006.04.004
    [7] G. H. Golub, C. F. V. Loan, Matrix computations, Baltimore: Johns Hopkins University Press, 2013. Available from: https://epubs.siam.org/doi/book/10.1137/1.9781421407944
    [8] K. Zuo, Y. Chen, L. Yuan, Further representations and computations of the generalized Moore-Penrose inverse, AIMS Math., 8 (2023), 23442–23458. https://doi.org/10.3934/math.20231191 doi: 10.3934/math.20231191
    [9] W. R. Hamilton, On quaternions, or on a new system of imaginaries in algebra, Philos. Mag., 25 (1844), 489–495. https://doi.org/10.1080/14786444408645047 doi: 10.1080/14786444408645047
    [10] S. D. Leo, G. Scolarici, Right eigenvalue equation in quaternionic quantum mechanics, J. Phys. A, 33 (2000), 2971–2995. http://doi.org/10.1088/0305-4470/33/15/306 doi: 10.1088/0305-4470/33/15/306
    [11] C. C. Took, D. P. Mandic, Augmented second-order statistics of quaternion random signals, Signal Process., 91 (2011), 214–224. https://doi.org/10.1016/j.sigpro.2010.06.024 doi: 10.1016/j.sigpro.2010.06.024
    [12] S. L. Adler, Quaternionic quantum mechanics and quantum fields, New York: Oxford University Press, 1995. Available from: https://onlinelibrary.wiley.com/doi/abs/10.1002/qua.560600402
    [13] J. B. Kuipers, Quaternions and rotation sequences, Princeton: Princeton University Press, 1999.
    [14] A. Rehman, I. I. Kyrchei, I. Ali, M. Akram, A. Shakoor, The general solution of quaternion matrix equation having η-skew-Hermicity and its Cramer's rule, Math. Probl. Eng., 2019 (2019), 7939238. https://doi.org/10.1155/2019/7939238 doi: 10.1155/2019/7939238
    [15] A. Rehman, I. I. Kyrchei, I. Ali, M. Akram, A. Shakoor, Explicit formulas and determinantal representation for η-skew-Hermitian solution to a system of quaternion matrix equations, Filomat, 34 (2020), 2601–2627. https://doi.org/10.2298/FIL2008601R doi: 10.2298/FIL2008601R
    [16] A. Rehman, I. I. Kyrchei, Solving and algorithm to system of quaternion Sylvester-Type matrix equations with -hermicity, Adv. Appl. Clifford Algebras, 32 (2022), 49. https://doi.org/10.1007/s00006-022-01222-2 doi: 10.1007/s00006-022-01222-2
    [17] Z. H. He, Q. W. Wang, Y. Zhang, A simultaneous decomposition for seven matrices with applications, J. Comput. Appl. Math., 349 (2019), 93–113. https://doi.org/10.1016/j.cam.2018.09.001 doi: 10.1016/j.cam.2018.09.001
    [18] S. W. Yu, Z. H. He, T. C. Qi, X. X. Wang, The equivalence canonical form of five quaternion matrices with applications to imaging and Sylvester-type equations, J. Comput. Appl. Math., 393 (2021), 113494. https://doi.org/10.1016/j.cam.2021.113494 doi: 10.1016/j.cam.2021.113494
    [19] E. K. W. Chu, L. Hou, D. B. Szyld, J. Zhou, Numerical solution of singular Sylvester equations, J. Comput. Appl. Math., 436 (2024), 115426. https://doi.org/10.1016/j.cam.2023.115426 doi: 10.1016/j.cam.2023.115426
    [20] X. Shao, Y. Wei, E. K. Chu, Numerical solutions of quaternionic Riccati equations, J. Appl. Math. Comput., 69 (2023), 2617–2639. https://doi.org/10.1007/s12190-023-01848-w doi: 10.1007/s12190-023-01848-w
    [21] L. S. Liu, S. Zhang, A coupled quaternion matrix equations with applications, J. Appl. Math. Comput., 69 (2023), 4069–4089. https://doi.org/10.1007/s12190-023-01916-1 doi: 10.1007/s12190-023-01916-1
    [22] Z. H. He, Some new results on a system of Sylvester-type quaternion matrix equations, Lin. Multilin. Algebra, 69 (2021), 3069–3091. https://doi.org/10.1080/03081087.2019.1704213 doi: 10.1080/03081087.2019.1704213
    [23] Z. H. He, X. X. Wang, Y. F. Zhao, Eigenvalues of quaternion tensors with applications to color video processing, J. Sci. Comput., 94 (2023). https://doi.org/10.1007/s10915-022-02058-5
    [24] Z. H. He, C. Navasca, X. X. Wang, Decomposition for a quaternion tensor triplet with applications, Adv. Appl. Clifford Algebras, 32 (2022), 9. https://doi.org/10.1007/s00006-021-01195-8 doi: 10.1007/s00006-021-01195-8
    [25] S. B. Aoun, N. Derbel, H. Jerbi, T. E. Simos, S. D. Mourtas, V. N. Katsikis, A quaternion Sylvester equation solver through noise-resilient zeroing neural networks with application to control the SFM chaotic system, AIMS Math., 8 (2023), 27376–27395. Available from: https://www.aimspress.com/article/doi/10.3934/math.20231401
    [26] M. Liu, H. Wu, Y. Shi, L. Jin, High-order robust discrete-time neural dynamics for time-varying multi-linear tensor equation with M-tensor, IEEE Trans. Ind. Inform., 9 (2023), 9457–9467. http://dx.doi.org/ 10.1109/TII.2022.3228394 doi: 10.1109/TII.2022.3228394
    [27] J. Respondek, Matrix black box algorithms-a survey, Bull. Pol. Acad. Sci. Tech. Sci., 2022, e140535. https://dx.doi.org/10.24425/bpasts.2022.140535
    [28] I. I. Kyrchei, Cramer's rule for quaternionic systems of linear equations, J. Math. Sci., 155 (2008), 839–858. https://doi.org/10.1007/s10958-008-9245-6 doi: 10.1007/s10958-008-9245-6
    [29] I. I. Kyrchei, The theory of the column and row determinants in a quaternion linear algebra, Adv. Math. Resear., 15 (2012), 301–359. Available from: https://www.elibrary.ru/item.asp?id=29685532
    [30] I. I. Kyrchei, Determinantal representations of the quaternion weighted Moore-Penrose inverse and its applications, Adv. Math. Resear., 23 (2017), 35–96. Available from: https://www.elibrary.ru/item.asp?id=35708733
    [31] I. I. Kyrchei, Determinantal representations of the Drazin and W-weighted Drazin inverses over the quaternion skew field with applications, Quater. Theory Appl., 2017,201–275. Available from: https://www.elibrary.ru/item.asp?id = 38610582
    [32] I. I. Kyrchei, Cramer's Rules of η-(skew-)Hermitian solutions to the quaternion Sylvester-type matrix equations, Adv. Appl. Clifford Algebras, 29 (2019), 56. https://doi.org/10.1007/s00006-019-0972-1 doi: 10.1007/s00006-019-0972-1
    [33] I. I. Kyrchei, Determinantal representations of solutions to systems of two-sided quaternion matrix equations, Lin. Multilin. Algebra, 69 (2021), 648–672. https://doi.org/10.1080/03081087.2019.1614517 doi: 10.1080/03081087.2019.1614517
    [34] I. I. Kyrchei, Determinantal representations of general and (skew-)Hermitian solutions to the generalized Sylvester-type quaternion matrix equation, Abstr. Appl. Anal., 2019 (2019), 5926832. https://doi.org/10.1155/2019/5926832 doi: 10.1155/2019/5926832
    [35] O. Alshammari, M. Kchaou, H. Jerbi, S. B. Aoun, V. Leiva, A fuzzy design for a sliding mode observer-based control scheme of Takagi-Sugeno Markov jump systems under imperfect premise matching with bio-economic and industrial applications, Mathematics, 10 (2022), 3309. https://doi.org/10.3390/math10183309 doi: 10.3390/math10183309
    [36] P. B. Dhandapani, J. Thippan, C. Martin-Barreiro, V. Leiva, C. Chesneau, Numerical solutions of a differential system considering a pure hybrid fuzzy neutral delay theory, Electronics, 11 (2022), 1478. https://doi.org/10.3390/electronics11091478 doi: 10.3390/electronics11091478
    [37] M. A. Akbar, V. Leiva, A new taxonomy of global software development best practices using prioritization based on a fuzzy system, J. Softw. Evol. Proc., 36 (2024). https://doi.org/10.1002/smr.2629
    [38] R. G. Aykroyd, V. Leiva, F. Ruggeri, Recent developments of control charts, identification of big data sources and future trends of current research, Technol. Forecast. Soc. Change, 144 (2019), 221–232. https://doi.org/10.1016/j.techfore.2019.01.005 doi: 10.1016/j.techfore.2019.01.005
    [39] A. Ghaffar, M. Z. Rahman, V. Leiva, C. Martin-Barreiro, X. Cabezas, C. Castro, Efficiency, optimality, and selection in a rigid actuation system with matching capabilities for an assistive robotic exoskeleton, Eng. Sci. Technol., 51 (2024), 101613. https://doi.org/10.1016/j.jestch.2023.101613 doi: 10.1016/j.jestch.2023.101613
    [40] A. Rehman, Q. W. Wang, Z. H. He, Solution to a system of real quaternion matrix equations encompassing η-Hermicity, Appl. Math. Comput., 265 (2015), 945–957. https://doi.org/10.1016/j.amc.2015.05.104 doi: 10.1016/j.amc.2015.05.104
    [41] A. Rehman, Q. W. Wang, I. Ali, M. Akram, M. O. Ahmad, A constraint system of generalized Sylvester quaternion matrix equations, Adv. Appl. Clifford Algebr., 3 (2017), 3183–3196. https://doi.org/10.1007/s00006-017-0803-1 doi: 10.1007/s00006-017-0803-1
    [42] A. Rehman, I. I. Kyrchei, I. Ali, M. Akram, A. Shakoor, Constraint solution of a classical system of quaternion matrix equations and its Cramer's rule, Iran J. Sci. Technol. Trans. Sci., 45 (2021), 1015–1024. https://doi.org/10.1007/s40995-021-01083-7 doi: 10.1007/s40995-021-01083-7
    [43] Z. Z. Bai, On Hermitian and skew-Hermitian splitting iteration methods for continuous Sylvester equations, J. Comput. Math., 29 (2011), 185–198. https://dx.doi.org/10.4208/jcm.1009-m3152 doi: 10.4208/jcm.1009-m3152
    [44] J. K. Baksalary, R. Kala, The matrix equation AXYB=C, Linear Algebra Appl., 25 (1979), 41–43. https://doi.org/10.1016/0024-3795(79)90004-1 doi: 10.1016/0024-3795(79)90004-1
    [45] W. E. Roth, The equations AXYB=C and AXXB=C in matrices, Proc. Amer. Math. Soc., 3 (1952), 392–396. https://doi.org/10.2307/2031890 doi: 10.2307/2031890
    [46] L. Wang, Q. W. Wang, Z. H. He, The common solution of some matrix equations, Algebra Coll., 23 (2016), 71–81. https://doi.org/10.1142/S1005386716000092 doi: 10.1142/S1005386716000092
    [47] Q. W. Wang, Z. H. He, Solvability conditions and general solution for the mixed Sylvester equations, Automatica, 49 (2013), 2713–2719. https://doi.org/10.1016/j.automatica.2013.06.009 doi: 10.1016/j.automatica.2013.06.009
    [48] S. G. Lee, Q. P. Vu, Simultaneous solutions of matrix equations and simultaneous equivalence of matrices, Lin. Alg. Appl., 437 (2012), 2325–2339. https://doi.org/10.1016/j.laa.2012.06.004 doi: 10.1016/j.laa.2012.06.004
    [49] Y. Q. Lin, Y. M. Wei, Condition numbers of the generalized Sylvester equation, IEEE Trans. Automat. Control, 52 (2007), 2380–2385. http://doi.org/10.1109/TAC.2007.910727 doi: 10.1109/TAC.2007.910727
    [50] X. Zhang, A system of generalized Sylvester quaternion matrix equations and its applications, Appl. Math. Comput., 273 (2016), 74–81. https://doi.org/10.1016/j.amc.2015.09.074 doi: 10.1016/j.amc.2015.09.074
    [51] Z. H. He, Q. W. Wang, A pair of mixed generalized Sylvester matrix equations, J. Shanghai Univ. Nat. Sci., 20 (2014), 138–156. http://doi.org/10.3969/j.issn.1007-2861.2014.01.021 doi: 10.3969/j.issn.1007-2861.2014.01.021
    [52] Q. W. Wang, A. Rehman, Z. H. He, Y. Zhang, Constraint generalized Sylvester matrix equations, Automatica, 69 (2016), 60–64. https://doi.org/10.1016/j.automatica.2016.02.024 doi: 10.1016/j.automatica.2016.02.024
    [53] F. O. Farid, Z. H. He, Q. W. Wang, The consistency and the exact solutions to a system of matrix equations, Lin. Multilin. Algebra, 64 (2016), 2133–2158. https://doi.org/10.1080/03081087.2016.1140717 doi: 10.1080/03081087.2016.1140717
    [54] Z. H. He, Q. W. Wang, A system of periodic discrete-time coupled Sylvester quaternion matrix equations, Algebra Coll., 24 (2017), 169–180. https://doi.org/10.1142/S1005386717000104 doi: 10.1142/S1005386717000104
    [55] X. Liu, Z. H. He, η-Hermitian solution to a system of quaternion matrix equations, Bull. Malaysian Math. Sci. Soc., 43 (2020), 4007–4027. https://doi.org/10.1007/s40840-020-00907-w doi: 10.1007/s40840-020-00907-w
    [56] Q. W. Wang, Z. H. He, Systems of coupled generalized Sylvester matrix equations, Automatica, 50 (2014), 2840–2844. https://doi.org/10.1016/j.automatica.2014.10.033 doi: 10.1016/j.automatica.2014.10.033
    [57] Z. H. He, A system of coupled quaternion matrix equations with seven unknowns and its applications, Adv. Appl. Clifford Algebras, 29 (2019), 38. https://doi.org/10.1007/s00006-019-0955-2 doi: 10.1007/s00006-019-0955-2
    [58] V. L. Syrmos, F. L. Lewis, Output feedback eigenstructure assignment using two Sylvester equations, IEEE Trans. Autom. Cont., 38 (1993), 495–499. http://doi.org/10.1109/9.210155 doi: 10.1109/9.210155
    [59] R. C. Li, A bound on the solution to a structured Sylvester equation with an application to relative perturbation theory, SIAM J. Matrix Anal. Appl., 21 (1999), 440–445. https://doi.org/10.1137/S0895479898349586 doi: 10.1137/S0895479898349586
    [60] G. Marsaglia, G. P. H. Styan, Equalities and inequalities for ranks of matrices, Lin. Multilin. Algebra, 2 (1974), 269–292. https://doi.org/10.1080/03081087408817070 doi: 10.1080/03081087408817070
    [61] Q. W. Wang, Z. C. Wu, C. Y. Lin, Extremal ranks of a quaternion matrix expression subject to consistent systems of quaternion matrix equations with applications, Appl. Math. Comput., 182 (2006), 1755–1764. https://doi.org/10.1016/j.amc.2006.06.012 doi: 10.1016/j.amc.2006.06.012
    [62] Z. H. He, Q. W. Wang, The general solutions to some systems of matrix equations, Lin. Multilin. Algebra, 63 (2015), 2017–2032. https://doi.org/10.1080/03081087.2014.896361 doi: 10.1080/03081087.2014.896361
    [63] I. I. Kyrchei, Determinantal representations of the Moore-Penrose inverse over the quaternion skew field and corresponding Cramer's rules, Lin. Multilin. Algebra, 59 (2011), 413–431. https://doi.org/10.1080/03081081003586860 doi: 10.1080/03081081003586860
    [64] Y. Zhang, J. Zhang, J. Weng, Dynamic Moore-Penrose inversion with unknown derivatives: Gradient neural network approach, IEEE Trans. Neur. Net. Learn. Syst., 34 (2023), 10919–10929. http://doi.org/10.1109/TNNLS.2022.3171715 doi: 10.1109/TNNLS.2022.3171715
    [65] Y. Zhang, Improved GNN method with finite-time convergence for time-varying Lyapunov equation, Inform. Sci., 611 (2022), 494–503. https://doi.org/10.1016/j.ins.2022.08.061 doi: 10.1016/j.ins.2022.08.061
  • This article has been cited by:

    1. Hesham A. Hefny, Nelly S. Amer, Interval Type-2 Mutual Subsethood Cauchy Fuzzy Neural Inference System (IT2MSCFuNIS), 2024, 17, 1875-6883, 10.1007/s44196-024-00405-y
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1121) PDF downloads(132) Cited by(3)

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog