A Fault Estimation and Fault-Tolerant Control Method Based on an Unknown-Input Sliding-Mode Observer

Abstract

The present invention provides a fault estimation and fault-tolerant control method based on an unknown input sliding mode observer, belonging to the technical field of fault-tolerant control of flexible manipulators. It solves the technical problem that the flexible manipulator system cannot operate stably in the face of unknown system states, actuator faults, and sensor faults. The technical solution is as follows: It includes the following steps: S1. Establish a Takagi-Sugeno fuzzy model; S2. Set the sensor fault type and establish an augmented system; S3. Develop an unknown input sliding mode observer to obtain a dynamic error system; S4. Accurately estimate and reconstruct faults and unknown states; S5. Develop an SMC measurement strategy to keep the flexible manipulator system operating stably when a fault occurs. The beneficial effects of the present invention are: The present invention can simultaneously estimate and reconstruct faults and unknown system states, and can restore the performance of the system in the face of unpredictable sensor faults and actuator faults.

Description

Technical Field

[0001] The present invention relates to the technical field of flexible manipulator control, and particularly to a fault estimation and fault-tolerant control method based on an unknown input sliding mode observer. Background Technique

[0002] With the rapid development of intelligent technology and industry, the application of manipulators in complex working environments has gradually increased. Traditional rigid manipulators and flexible manipulators play important roles in tasks such as handling and grasping. Flexible manipulators have attracted much attention due to their advantages such as fast response, lightweight structure, low energy consumption, and high load-to-self-weight ratio. However, due to the characteristics of flexible structures, flexible manipulators are often accompanied by elastic vibrations during movement, which increases the difficulty of control design. Compared with rigid manipulators, flexible manipulators tend to generate elastic vibrations during operation due to their structures and materials having the characteristic of low bending stiffness, increasing the complexity of the motion state of flexible manipulators and making the control design of flexible manipulators more difficult. At the same time, with the increasing complexity of actual engineering systems, the failure rate of control systems is also getting higher and higher. Due to the characteristics of flexible materials, more complex sensors and actuators are often required to achieve precise control. Once a sensor, actuator or component in the flexible manipulator system fails, it will seriously affect the accuracy and stability of the entire closed-loop system. Therefore, it is particularly important to study the fault-tolerant control method for actuators and sensors of flexible manipulators.

[0003] To solve the harm caused by unpredictable faults, a series of achievements have been made in fault diagnosis (FD), fault reconstruction (FR), and fault-tolerant control (FTC) to ensure the safety and reliability of engineering systems. In recent years, FTC design methods for complex nonlinear systems, stochastic systems, and Markov jump systems have also been widely studied. The literature "Fault reconstruction and fault-tolerant control via learning observers in Takagi–Sugeno fuzzy descriptor systems with time delays" (Q. Jia, W. Chen, Y. Zhang, and H. Li, IEEE Transactions on Industrial Electronics, vol. 62, no. 6, pp. 3885-3895, 2015.) estimates the unknown system state and actuator faults through a designed fuzzy learning observer, and ensures the reliability of the control system under actuator faults through a fuzzy FTC scheme. Although it has achieved results in the control of disturbed unmanned aerial vehicle systems with uncertain parameters, this control method has limited fault tolerance and poor adaptability, and is not suitable for actual operating environments with changing disturbance amounts.

[0004] From the theoretical perspective of sliding mode variable structure control, the sliding mode variable structure control scheme has received extensive attention and research due to its advantages such as simple design, fast response speed, and strong robustness, and is often used in the control of complex flexible joint manipulators. The literature "Adaptive speed control of PMSM drive system based a new sliding-mode reaching law" (A.K. Junejo, W. Xu, C. Mu, M.M. Ismail, and Y. Liu, IEEE Transactions on Power Electronics, vol. 35, no. 11, pp. 12110-12121, 2020.) proposed the terminal sliding mode control method, boundary layer method, reaching law method, etc. Among them, compared with the terminal SMC and boundary layer methods, the reaching law is simple in design and does not reduce the system performance, and has been widely applied and studied. At present, although several methods of reaching law have been proposed to reduce chattering, the problem of system chattering is still inevitable. How to improve the chattering suppression ability of the system remains a challenging topic. Summary of the Invention

[0005] Aiming at the technical problem of fault-tolerant control of a flexible manipulator system in the case of simultaneous occurrence of unpredictable sensor faults and actuator faults, the purpose of the present invention is to provide a fault estimation and fault-tolerant control method based on an unknown input sliding mode observer, which can simultaneously estimate and reconstruct faults and unknown system states, and can make the system recover its performance when facing unpredictable sensor faults and actuator faults, significantly improving the reliability of the flexible manipulator sliding mode control system when a fault occurs; at the same time, by reducing the chattering of the control system, the operation stability of the system is further optimized.

[0006] The inventive concept of the present invention is: The present invention provides a fault estimation and fault-tolerant control method based on an unknown input sliding mode observer, which specifically includes the following steps:

[0007] S1. According to the working characteristics of the flexible joint manipulator system, considering the possible external non-linear interference situation, construct a Takagi-Sugeno fuzzy system describing sensor faults and actuator faults; S2. Based on the designed Takagi-Sugeno fuzzy system with sensor faults and actuator faults, set the sensor faults to be independent of time to obtain an augmented system;

[0008] S3. Based on the designed augmented system, an unknown input sliding mode observer (UISMO) was developed to obtain the dynamic error system; S4. The input signal of the UISMO based on the integral switching function was designed, and the augmented error system was constructed through the UISMO to ensure the stability of the system and achieve accurate estimation and reconstruction of faults and unknown states; S5. Based on the fault information measured by the unknown input sliding mode observer with the integral switching function, a sliding mode fault-tolerant control method (SMFTC) was designed to ensure the stable operation of the flexible manipulator system during faults and further improve the stability of the flexible manipulator sliding mode control system under fault conditions.

[0009] To achieve the above invention purpose, the technical solution adopted by the present invention is specifically as follows: To achieve the above invention purpose, the technical solution adopted by the present invention is specifically as follows:

[0010] S1. According to the working characteristics of the flexible joint manipulator system and considering the possible external non-linear interference, a Takagi-Sugeno fuzzy system describing sensor faults and actuator faults was constructed.

[0011] S2. Based on the Takagi-Sugeno fuzzy system designed in step S1 to describe sensor faults and actuator faults, assuming that the sensor fault is independent of time, an augmented system was obtained.

[0012] S3. Based on the augmented system designed in step S2, in order to obtain the information of faults and unknown states simultaneously, an unknown input sliding mode observer was developed to obtain the dynamic error system.

[0013] S4. An augmented error system was constructed through the UISMO, and the input signal of the UISMO based on the integral switching function and the adaptive reaching law ARL were designed to ensure the stability of the system and achieve accurate estimation and reconstruction of faults and unknown states.

[0014] S5. Based on the unknown input sliding mode observer with the integral switching function designed in step S3 and step S4, a sliding mode fault-tolerant control method was designed according to the measured fault information, so that the flexible manipulator system can maintain stable operation during faults and further improve the stability of the flexible manipulator sliding mode control system under fault conditions.

[0015] Furthermore, as a preferred technical solution of the present invention, S1 includes the following steps:

[0016] According to the working characteristics of the flexible joint manipulator system, a clear structural model of the flexible joint manipulator system was established:

[0017]

[0018] In formula (1), J land J m are the motor inertia and the link inertia respectively; θ l and θ m represent the angular positions of the link and the motor respectively; M is the mass of the link; l is the distance from the link center to the joint axis; K and F represent the joint spring stiffness coefficient and the rotor friction coefficient respectively; g is the gravitational constant. Among them, the external disturbance is expressed as Let be the state variables, and the above flexible joint manipulator system can be reconstructed as:

[0019]

[0020] In Equation (2),

[0021] Among them,

[0022] Based on the nonlinear system with external disturbances, a T-S fuzzy model describing sensor faults and actuator faults is constructed. The form of this model described by the IF-THEN framework is as follows:

[0023] Rule i: If z1 is is then

[0024]

[0025] In Equation (3), the system state is defined as The input and output are defined as and f(t) is the external disturbance and satisfies f(t) ≤ α + β||x(t)||; the actuator fault is defined as The sensor fault is defined as and satisfies A i , B, C, D and E are matrices with appropriate dimensions, where C and E satisfy rank(CE) = rank(E), then the T-S model can be further derived as:

[0026]

[0027] Among them, z(t) = [z1(t)…z p (t)] T ; θ i (z(t)) is the membership function in fuzzy rule i, and the form is as follows:

[0028]

[0029] In Equation (5), is the antecedent variable zj (t) Membership degree in the fuzzy set ; θ i (z(t)) satisfies 1≥θ i (z(t))≥0, and

[0030] Furthermore, as a preferred technical solution of the present invention, the S2 includes the following steps:

[0031] Assume that the sensor fault is independent of time, that is The obtained augmented system (2) is as follows:

[0032]

[0033] In formula (6), x a =[x(t) f s (t)] T , C a =[C D],

[0034] Furthermore, as a preferred technical solution of the present invention, the S3 includes the following steps:

[0035] To simultaneously achieve accurate estimation of unknown system states, sensor faults, and actuator faults, consider the following unknown input sliding mode observer (UISMO):

[0036]

[0037] In formula (7), C E =(C a E a ) + is the generalized inverse matrix of matrix C a E a ; and are the state and input signals of the observer (7) respectively; and represent the actuator estimation signal and state estimation signal of the augmented system (2) respectively; G, H i , L i , M1 and N i are a set of gains to be designed; the output derivative

[0038] is the actuator fault estimation value calculated from the continuous-time output signal y(t). Define the following dynamic error form:

[0039]

[0040] Among them, the state error e of the systemxa (t) is:

[0041]

[0042] In Equation (9), The error dynamics are described as follows:

[0043]

[0044] Define the parameter matrix in Equation (7) and eliminate the unknown term f a (t) and u(t)

[0045] FE a = 0 (11)

[0046] M1 = FB a (12)

[0047] N i = L i Q - R i (13)

[0048] L i = FA ai + R i C a (14)

[0049] The error dynamics are further expressed as:

[0050]

[0051] The error system described in Equation (10) is independent of the actuator fault. In addition, the error of the actuator fault can be expressed as:

[0052]

[0053] Based on C E C a E a = (C a E a ) + C a E a = I, select H i = C E C a A ai = C E C a B a The error of the actuator fault can be further expressed as:

[0054]

[0055] The input signal v(t) of the designed UISMO enables e xa (t) → 0 and the actuator fault estimation error e fa (t) → 0 to achieve synchronization, so as to ensure the subsequent augmented singular system remains stable.

[0056]

[0057] Among them, Considering FE a = 0, rank(C a E a ) = rank(CE) = rank(E), then there is E a = QC a E a , and the matrix Q can be calculated by the following formula:

[0058] Q = φ1-Tφ2 (19)

[0059] Among them, φ1 = E a (C a E a ) + , T is an arbitrary matrix to be designed, and then it can be obtained:

[0060]

[0061] Furthermore, as a preferred technical solution of the present invention, the S4 includes the following steps:

[0062] Design the input signal v(t) to make the augmented singular system (18) converge to zero, and select the following integral switching function:

[0063]

[0064] In formula (21), the value of G e is determined by the constraint conditions of the operation of the control system in the actual situation, that is is an invertible matrix. The derivative of the switching function s e (t) is in the following form:

[0065]

[0066] According to the SMC theory, obtain the equivalent control law:

[0067] v eg (t) = f(t) (23)

[0068] Combining the adaptive reaching law ARL and the equivalent control law v eq(t), the input signal v(t) in UISMO can be designed as:

[0069]

[0070] To reduce system chattering, the designed ARL adopts the following damped sine function:

[0071]

[0072] where 0 < λ1, λ2, λ3 < 1 and γ1 > 1. For the given augmented error system (5), under the SMC input signal v(t) in the above (24), the following Lyapunov function is constructed:

[0073]

[0074] Thus, it is proved that the state trajectory can reach the sliding mode surface S1 = {e xa (t): s e (t) = 0}. When the augmented singular system reaches the sliding mode surface, the system has the following sliding mode dynamics:

[0075]

[0076] For the sliding mode dynamics (27) of the known matrices A ai , C a , equations (11)-(14) and equation (20), if matrices P1 > 0, X1, X2, Y 1i , Y 2i , P2 and P3 can be found to satisfy the following inequalities, then the system state trajectory (27) is asymptotically stable:

[0077]

[0078] where Φ1 and Φ2 are defined by equation (19);

[0079] The matrix T in equation (19) and the matrix R in equation (13) i can be calculated respectively as T = P -1 X and R i = P -1 Y i , thus UISMO and its input signal v(t) can be designed.

[0080] By calculating the parameters in UISMO and synthesizing UISMO with the input signal v(t) designed in (15), the estimated signals and x a (t) can be generated, thereby reconstructing the system state and sensor faults in the augmented system (2) into their estimated values.

[0081]

[0082] Furthermore, as a preferred technical solution of the present invention, a sliding mode fault-tolerant control method (SMFTC) is designed according to the fault information measured by UISMO. The S5 includes the following steps:

[0083] Define and give an integral switching function in the following form:

[0084]

[0085] where, G c is a designable matrix and satisfies that the matrix G c B is invertible; and are the estimates of x(t) and f a (t) by UISMO respectively. The switching function (t) is differentiated with respect to time:

[0086]

[0087] According to the SMC theory, the following equivalent control scheme is obtained:

[0088]

[0089] In formula (18), is consistent with the designed input signal v(t) in formula (15). The SMC scheme u(t) is constructed as follows:

[0090]

[0091] Under the SMC scheme (32), for the states in system (4), the following Lyapunov function is constructed:

[0092]

[0093] The reachability of the sliding surface S1 = {x(t): s x (t) = 0} is proven, and the following sliding mode dynamics is obtained:

[0094]

[0095] To overcome 's influence, assuming rank(BE) = rank(B), the generalized inverse form of B is obtained as follows:

[0096] K a =-B + E (35)

[0097] Thus, BKa +E = 0. Consider The sliding mode dynamic can be reconfigured as:

[0098]

[0099] Combining Equation (27) and Equation (34), the following augmented system is derived:

[0100]

[0101] where, Δ 1i = A i + BK, Δ 3i = FA ai + R i C a + M1K oi . Using the defined matrices F, M1, G c , H i , K oi Find the control gain to ensure the stability of the augmented system (37).

[0102] In addition, given the system matrices A i , B, E, A ai , C a , find the matrices and to make the inequality (38) hold, which can ensure the asymptotic stability of the augmented sliding mode dynamic system (37).

[0103]

[0104] Thus, the gain of SMC is obtained Combining the sliding mode control scheme based on UISMO designed in Equation (31), a fault compensation strategy can be generated.

[0105] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0106] 1. The present invention first proposes a fault estimation and sliding mode fault-tolerant control method for a flexible joint manipulator based on an unknown input sliding mode observer, which improves the reliability of the sliding mode control system of the flexible manipulator when a fault occurs, reduces the chattering of the control system, and realizes the high-robustness motion control of the flexible manipulator control system.

[0107] 2. Considering the characteristics of the actual working environment of the complex flexible manipulator, a new type of unknown input sliding mode observer is designed, which can simultaneously estimate the unknown states and faults of the system; this method can not only handle faults but also effectively handle external disturbances, improving the robustness of the system.

[0108] 3. An unknown input sliding mode observer with an integral switching function is introduced to achieve the joint estimation and reconstruction of faults and unknown states, breaking through the limitation that the actuator fault signal must be a step signal and enhancing the applicability of the system.

[0109] 4. Combining with an adaptive reaching law using a damped sine function, an observer-based sliding mode fault-tolerant control scheme is designed to reduce the chattering of the Takagi-Sugeno fuzzy control system; compared with the traditional reaching law, the chattering of this system is significantly reduced, and it can improve the anti-disturbance and error handling performance of the control system, thereby further enhancing the self-adaptability of the control system. Description of the Drawings

[0110] The drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation to the present invention.

[0111] Figure 1 It is a schematic diagram of the fault estimation and sliding mode fault-tolerant control method for a flexible-joint manipulator based on an unknown input sliding mode observer of the present invention.

[0112] Figure 2 It is a structural diagram of the flexible manipulator of the present invention.

[0113] Figure 3 It is a curve diagram of the sliding mode fault-tolerant control input signal and the observer input signal of the flexible manipulator of the present invention.

[0114] Figure 4 It is a curve diagram of the state trajectory of the unknown input observer of the present invention.

[0115] Figure 5 It is a curve diagram of the link angle position, link angular velocity and their estimates of the flexible manipulator of the present invention.

[0116] Figure 6 It is a curve diagram of the motor angle position, motor angular velocity and their estimates of the flexible manipulator of the present invention.

[0117] Figure 7 It is a curve diagram of the estimation errors of the link angle, angular velocity and motor angle angular velocity of the flexible manipulator of the present invention.

[0118] Figure 8 It is a curve diagram of the control system input of the flexible manipulator of the present invention.

[0119] Figure 9 It is a curve diagram of the actuator and sensor faults and their estimates of the present invention.

[0120] Figure 10 It is a curve diagram of the estimation errors of the actuator and sensor faults of the present invention.

[0121] Figure 11 The state trajectory curve diagram of the flexible manipulator system under the adaptive reaching law and the traditional reaching law proposed by the present invention.

[0122] Figure 12 The input curve diagram of the flexible manipulator system under the adaptive reaching law and the traditional reaching law proposed by the present invention. Specific implementation manners

[0123] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Of course, the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0124] Embodiment 1

[0125] Refer to Figures 1 to 12 , this embodiment provides its technical solution as a fault estimation and fault-tolerant control method based on an unknown input sliding mode observer. Please refer to Figure 1 , which is the schematic diagram of the fault estimation and sliding mode fault-tolerant control method for a flexible joint manipulator based on an unknown input sliding mode observer provided by the present invention, as shown in Figure 1 shown.

[0126] The structure diagram of the flexible manipulator is as shown in Figure 2 shown. According to the working characteristics of the flexible joint manipulator system, a clear structural model of the flexible joint manipulator system is established:

[0127]

[0128] In formula (1), J l and J m are the motor inertia and the link inertia respectively; θ l and θ m θ m represent the angular positions of the link and the motor respectively; M is the mass of the link; l is the distance from the center of the link to the joint axis; K and F represent the joint spring stiffness coefficient and the rotor friction coefficient respectively; g is the gravitational constant. Among them, the external disturbance is expressed as Let be the state variables, and the above flexible joint manipulator system can be reconstructed:

[0129]

[0130] In formula (2),

[0131] Among them, Considering Combined with actuator faults and sensor faults, the nonlinear system (2) described by the IF-THEN framework is in the following form:

[0132] Rule1: IF x1(t) is about 0, THEN

[0133]

[0134] Rule2: IF x1(t) is about THEN

[0135]

[0136] Where

[0137]

[0138] Let the system parameters be M = 0.02 kg, l = 1 m, J1 = J m = 1 kg·m 2 ², g = 9.81 m / s 2 ², F = 0.008 N·m·s / rad. Therefore, the overall T-S fuzzy model can be written in the following form:

[0139]

[0140] Where

[0141]

[0142] B = [0 0 0 1] T

[0143] The actuator and sensor fault assumptions are as follows:

[0144]

[0145] is the membership function.

[0146] To solve the problem of the unmeasurability of the system state and the unpredictable faults occurring in the actuators and sensors, an unknown input sliding mode observer (UISMO) is first designed. By solving the inequality (28), the observer gain obtained is calculated as follows:

[0147]

[0148]

[0149] H1 = [-0.1962 1 0 0.1962 0]

[0150] H2 = [-0.1249 1 0 0.1962 0]

[0151] M1 = [0 0 0 0 0] T , G = 1

[0152] C E = [0 0 0 1]

[0153] Secondly, by solving the inequality (38), the sliding mode control gains obtained are as follows:

[0154] K c1 = [0.4395 -2.8417 -1.7407 -1.4832]

[0155] K c2 = [0.4551 -3.2075 -1.8793 -1.5652]

[0156] K a = -1

[0157] The remaining parameters are selected as α = 2, β = 2, λ1 = 0.7, λ2 = 0.1, λ3 = 0.5, G e = [0 0 0 0 -1], γ1 = 10, G c = [0 0 0 1]; The initial conditions are x(0) = [1 -1 π / 4 -1 / 2], x o (0) = [0 0 0 0].

[0158] In Figure 3 shows the SMC input signal (32) of the original system and the SMC input signal (24) of the observer, while Figure 4 presents the state trajectory curve of the UISMO response. After adopting the designed SMC and UISMO schemes, Figure 5 and Figure 6 show the state trajectory of the original system (41) and its estimated curve. The estimation error and the system output signal are shown in Figure 7 and Figure 8 . Specifically, Figures 5 to 7 clearly shows that applying the designed UISMO can effectively estimate the unknown system states. In addition, the estimation curves of the actuator and sensor faults that occur in the system (41) for UISMO are shown in Figure 10 , and the corresponding estimation errors are shown in Figure 11 , which proves the effectiveness of UISMO in fault estimation. In addition, Figures 5 to 7 as well as Figure 10 confirm that all augmented system singular states e A (t) = [e x1(t) e x2 (t) e x3 (t) e x4 (t) e fs (t) e fa (t)] and the system state x(t) converges to the equilibrium point, thus verifying the effectiveness of the SMC scheme designed by the present invention.

[0159] To verify the superiority of the reaching law (25) designed by the present invention, replace the slaw in Equation (25) with a constant rate reaching law slaw = -sign(s(t)), and the reconstructed SMC scheme is as follows:

[0160]

[0161] The control input u(t) in Equations (34) and (54) and the system state trajectory x4(t) under the SMC schemes (34) and (54) are as Figure 11 and Figure 12 shown. The reaching law proposed by the present invention effectively suppresses the chattering phenomenon of the system state and the input signal, thus verifying the superiority of the reaching law proposed by the present invention.

[0162] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A fault estimation and fault-tolerant control method based on an unknown input sliding mode observer, characterized in that, It includes the following steps: S1. According to the working characteristics of the flexible joint manipulator system and the external non-linear disturbance situation, construct a Takagi-Sugeno fuzzy system that describes sensor faults and actuator faults; The Takagi-Sugeno fuzzy system: Among them, the system state is defined as θ l and θ m represent the angular positions of the connecting rod and the motor respectively. The input and output are defined as u(t) and y(t); f(t) is an external disturbance and satisfies ||f(t)|| ≤ α + β||x(t)||; the actuator fault is defined as f a (t), and the sensor fault is defined as f s (t); A i , B, C, D, and E are system matrices with appropriate dimensions, satisfying rank(CE) = rank(E); r = 2 is the number of fuzzy rules, z(t) = [z1(t) … z p (t)] T ; is the membership function in fuzzy rule i, where is the antecedent variable z j (t) in the fuzzy set ; θ i (z(t)) satisfies 1 ≥ θ i (z(t)) ≥ 0, and The system matrices are specifically represented as follows: where J l and J m are the inertia of the motor and the inertia of the connecting rod respectively; M is the mass of the connecting rod; l is the distance from the center of the connecting rod to the joint axis; K and F represent the joint spring stiffness coefficient and the rotor friction coefficient respectively; g is the gravitational constant; S2. Based on the Takagi-Sugeno fuzzy system designed in step S1 that describes sensor faults and actuator faults, assume that the sensor faults are independent of time to obtain an augmented system; S3. Based on the augmented system designed in step S2, in order to simultaneously obtain information on faults and unknown states, develop an unknown input sliding mode observer to obtain a dynamic error system; S4. Construct an augmented error system through the unknown input sliding mode observer, design the input signal and the adaptive reaching law ARL of the unknown input sliding mode observer based on the integral switching function to ensure the stability of the system and achieve accurate estimation and reconstruction of faults and unknown states; S5. Based on the unknown input sliding mode observer with an integral switching function designed in steps S3 and S4, design a sliding mode fault-tolerant control method according to the measured fault information to enable the flexible manipulator system to operate stably when a fault occurs.

2. The fault estimation and fault-tolerant control method based on an unknown input sliding mode observer according to claim 1, characterized in that, Step S2 includes the following steps: Based on the Takagi-Sugeno fuzzy system, it is assumed that the sensor fault is independent of time, that is The augmented system (2) is obtained as follows: where x a = [x(t) f s (t)] T , C a = [C D], 3. The fault estimation and fault-tolerant control method based on an unknown input sliding mode observer according to claim 2, characterized in that, Step S3 includes the following steps: To simultaneously achieve accurate estimation of unknown system states, sensor faults, and actuator faults, design the following unknown input sliding mode observer: where C E =(C a E a ) + is the generalized inverse matrix of matrix C a E a ; x o (t) and v(t) are the state and input signals of the unknown input sliding mode observer (3), respectively; and represent the actuator estimation signal and the state estimation signal of the augmented system (2), respectively; G, H i , L i , M1 and N i are a set of gains to be designed; the output derivative is the actuator fault estimation value calculated from the continuous-time output signal y(t), and the error dynamics is defined as where e xa (t) is the state error of the system, and the error dynamics is further expressed as: and the error system is independent of the actuator fault, based on C E C a E a =(C a E a ) + C a E a =I, select H i =C E C a A ai =C E C a B a , the actuator fault error is 4. The fault estimation and fault-tolerant control method based on an unknown input sliding mode observer according to claim 3, characterized in that, Step S4 includes the following steps: Design the input signal v(t) synchronization based on an unknown input sliding mode observer to make e xa (t) → 0 and the actuator fault estimation error e fa (t) → 0 to ensure the stability of the following augmented error system; Among them, In order to reduce the influence of chattering in the sliding mode control process on the system stability, the following adaptive reaching law ARL is designed: According to the SMC theory, the compensation input signal v(t) in the unknown input sliding mode observer is designed as: Among them, is an integral switching function designed for the augmented error system (5), where 0 < λ1, λ2, λ3 < 1, γ1 > 1, and the value of G e > 1 is determined by the constraint conditions of the operation of the control system in the actual situation, that is is an invertible matrix. By calculating the parameters in the unknown input sliding mode observer, the unknown input sliding mode observer is synthesized with the input signal v(t) designed in formula (7) to generate the estimated signals and so as to reconstruct the system state and sensor faults in the augmented system (2) into their estimated values 5. The fault estimation and fault-tolerant control method based on an unknown input sliding mode observer according to claim 4, characterized in that: Step S5 includes the following steps: Definition wherein According to the sliding mode control theory, the following fault-tolerant SMC control scheme based on ARL is designed: Among them, represents an integral switching function designed for the augmented system (2), where G c is the design matrix and satisfies that the matrix G c B is invertible.

Images