Intelligent chassis double-disturbance observer control method

Abstract

The invention discloses an intelligent chassis double-disturbance observer control method, and belongs to the technical field of disturbance control, and the method comprises the following steps: S1, dividing a control system into a non-matching disturbance subsystem and a matching disturbance subsystem; s2, performing disturbance compensation and disturbance estimation on the split matching disturbance subsystem by using an IUDE method to obtain an equivalent system; s3, determining disturbance boundaries of the equivalent system; and S4, using the weight matrix of state control and the weight matrix of disturbance estimation to perform disturbance estimation and compensation on the equivalent system after disturbance boundaries are determined. The method is applied to ride comfort control of an active suspension 1 / 4 vehicle model, the IUDE algorithm and the LMID algorithm effectively achieve estimation of original disturbance and equivalent disturbance respectively, and it is indicated that the method is well achieved in the two aspects of stabilizing a control system and achieving disturbance estimation.

Description

Technical Field

[0001] This invention belongs to the field of disturbance control technology, specifically relating to a smart chassis dual disturbance observer control method. Background Technology

[0002] Dynamic control of intelligent chassis systems typically refers to the use of different control algorithms to control the transmission system, driving system, steering system, braking system, and suspension system, thereby achieving intelligent chassis control.

[0003] Since these systems typically include a certain degree of disturbance, which acts on the dynamic system in the form of force or torque, the control of the system needs to possess a certain degree of robustness. Generally, the robustness of the system can be improved in two ways: first, by employing robust control algorithms to suppress disturbances; and second, by employing disturbance observer algorithms to estimate and compensate for disturbances.

[0004] However, robust control algorithms can only mitigate disturbances to a certain extent, and disturbance observer algorithms inevitably introduce errors during the estimation process. In other words, both schemes have certain shortcomings in disturbance suppression, which affects the improvement of key vehicle performance. Summary of the Invention

[0005] To address the above problems, this invention proposes a smart chassis dual-disturbance observer control method.

[0006] The technical solution of this invention is: a smart chassis dual-disturbance observer control method comprising the following steps:

[0007] S1. The control system is divided into a non-matched disturbance subsystem and a matched disturbance subsystem;

[0008] S2. The IUDE method is used to perform perturbation compensation and perturbation estimation on the split matched perturbation subsystem to obtain the equivalent system;

[0009] S3. Determine the boundedness of the disturbance of the equivalent system;

[0010] S4. Using the weight matrix of state control and the weight matrix of disturbance estimation, perform disturbance estimation and compensation on the equivalent system after determining the boundedness of the disturbance.

[0011] Furthermore, in S1, the expression for separating the unmatched perturbation subsystem and the matched perturbation subsystem is:

[0012] ;

[0013] in, Represents the state of the non-matched disturbance subsystem The derivative of Indicates the state of the matched disturbance subsystem The derivative of This represents the steady system matrix of the unmatched perturbation subsystem. This represents the steady system matrix of the matched perturbation subsystem. Represents the system's state variables. This represents the steady-state input matrix of the unmatched perturbation subsystem. This represents the steady-state input matrix of the matched perturbation subsystem. Indicates the system's control input, This represents the known terms of the non-matched perturbation subsystem. This represents the known terms of the matched perturbation subsystem. This indicates a non-matching perturbation. This indicates a matching perturbation.

[0014] Furthermore, S2 includes the following sub-steps:

[0015] S21. Perform disturbance compensation on the split matching disturbance subsystem;

[0016] S22. Perform disturbance estimation on the matched disturbance subsystem after disturbance compensation;

[0017] S23. Use the control quantity to perform equivalent processing on the matched disturbance subsystem after disturbance compensation and disturbance estimation to obtain the equivalent system.

[0018] Furthermore, in S21, the expression for perturbation compensation of the split matched perturbation subsystem is:

[0019] ;

[0020] in, This represents the steady system matrix of the matched perturbation subsystem. Represents the system's state variables. This represents the steady-state input matrix of the matched perturbation subsystem. Indicates reference input. This represents the known terms of the matched perturbation subsystem. This indicates a matching perturbation. This represents the compensation value for matching perturbations. Indicates the system's control input;

[0021] In S22, the discretized expression for perturbation estimation is:

[0022] ;

[0023] in, Indicates the matching perturbation exist The estimated value at time, Indicates to the first process quantity, represents the estimation value of the matching disturbance , represents the discrete time, represents the second process quantity, the first process quantity, represents the third process quantity the first process quantity, at the time, represents the third process quantity the first process quantity, represents the third process quantity at the time; In S23, the expression of the control quantity

[0024] is:

[0025]

[0026]

[0027] wherein, represents the reference control, represents the pseudo-inverse, represents the compensation value of the matching disturbance, , is the partial disturbance compensation ratio, represents the simple expression form of the diagonal matrix.

[0028] Further, in S3, the expression of the equivalent system is:

[0029]

[0030] wherein, represents the derivative of the state variable of the control system, represents the steady system matrix of the control system, represents the state variable of the control system, represents the steady input matrix of the control system, represents the reference control, represents the known item of the equivalent system, represents the disturbance of the equivalent system.

[0031] Further, in S3, when the original disturbance is bounded and the derivative of the matching disturbance is bounded, then the disturbance of the equivalent system is bounded.

[0032] Further, in S41, the expression of the weight matrix of the state control is:

[0033] ​​​​ ;

[0034] wherein, represents a simple expression form of a diagonal matrix, represents a state variable a weight of a component of the first dimension, represents a state variable a weight of a component of the first dimension, a weight of a component of the first dimension, represents a real number field, represents a state variable a dimension of the state variable.

[0035] In S41, the expression of the weight matrix of the disturbance estimation is as follows:

[0036] ;

[0037] wherein, represents a disturbance a weight of a component of the first dimension, represents a disturbance a weight of a component of the first dimension, a weight of a component of the first dimension, represents a simple expression form of a diagonal matrix.

[0038] The beneficial effects of the present application are:

[0039] (1) The LMID-IUDE algorithm of the present application simultaneously uses two kinds of ideas of a robust control algorithm (LMID algorithm) and a disturbance observer algorithm (IUDE algorithm); the IUDE algorithm is adopted to firstly process the matching disturbance of the original system and obtain an equivalent system and a disturbance; then the LMID algorithm is adopted to control the equivalent system and simultaneously obtain a disturbance observer, which can realize compensation of the matching disturbance of the equivalent system.

[0040] (2) The present application extends the related theory of the IUDE algorithm, which is embodied in two aspects, one is to give a discrete expression form of the disturbance estimation for facilitating practical application, and the other is to explain the boundedness of the disturbance of the equivalent system obtained by the algorithm.

[0041] (3) The present application extends the related theory of the LMID algorithm, introduces two weight matrices, which respectively correspond to the control of the system state and the estimation of the disturbance of the equivalent system, and enhances the flexibility of the algorithm; at the same time, since the LMID algorithm processes the equivalent system, the disturbance observer in the LMID algorithm can estimate the disturbance of the equivalent system and compensate the matching disturbance of the equivalent system.

[0042] (4) The application is applied to the smoothness control of the active suspension 1 / 4 vehicle model, the LMID-IUDE algorithm can effectively reduce the sprung mass acceleration compared with the IUDE algorithm or the LMID algorithm, thereby improving the smoothness, and the IUDE algorithm and the LMID algorithm effectively realize the estimation of the original disturbance and the equivalent disturbance respectively, which shows that the method has good realization in making the control system stable and realizing disturbance estimation. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 It is a flow chart of the intelligent chassis double-disturbance observer control method.

[0044] Figure 2 It is the structure of the designed LMID-IUDE control algorithm.

[0045] Figure 3 It is the active suspension 1 / 4 vehicle model.

[0046] Figure 4 It is a comparison control effect diagram of the designed control system on the random road, three performance indicators of the sprung mass acceleration, the wheel dynamic deformation and the suspension dynamic deflection.

[0047] Figure 5 It is a comparison diagram of the observation effect of the IUDE algorithm on the matching disturbance and the observation effect of the LMID algorithm on the equivalent matching disturbance. DETAILED DESCRIPTION

[0048] The embodiments of the application will be further described below with reference to the drawings.

[0049] As shown in the drawings, Figure 1 The application provides an intelligent chassis double-disturbance observer control method, which comprises the following steps:

[0050] S1, the control system is divided into a non-matching disturbance subsystem and a matching disturbance subsystem;

[0051] S2, the IUDE method is used for disturbance compensation and disturbance estimation on the divided matching disturbance subsystem to obtain an equivalent system;

[0052] S3, the disturbance boundedness of the equivalent system is determined;

[0053] S4, the disturbance estimation and compensation are performed on the equivalent system after the disturbance boundedness is determined by using the weight matrix of state control and the weight matrix of disturbance estimation.

[0054] The application simultaneously uses the robustness control algorithm (LMID algorithm) and the disturbance observer algorithm (IUDE algorithm) to enhance the robustness of the system, and the LMID algorithm additionally has the effect of the disturbance observer, and the whole includes two disturbance observers.

[0055] The application applies the IUDE algorithm to estimate and compensate the matching disturbance, gives a discrete form of the algorithm in practical application, and analyzes the boundedness of the equivalent system disturbance.

[0056] The application applies the LMID algorithm to control the equivalent system obtained by the IUDE algorithm, and realizes the estimation of the equivalent system disturbance. Meanwhile, the algorithm is introduced with weights to realize the flexible adjustment of state control and disturbance estimation.

[0057] The controller structure designed in the application is shown in Figure 2 The algorithm is named as LMID-IUDE algorithm, and includes the following four contents.

[0058] Firstly, the original system is split according to the non-matching disturbance and the matching disturbance.

[0059] Secondly, the IUDE algorithm is used to control the matching disturbance subsystem of the original system, to realize the estimation and compensation of the matching disturbance, and to obtain the discrete expression of the matching disturbance estimation as for the development of the actual controller.

[0060] Thirdly, the equivalent system obtained by the IUDE algorithm is analyzed, the matching disturbance compensation of the IUDE algorithm is expressed as , and the boundedness of the equivalent system disturbance is analyzed.

[0061] Fourthly, the LMID algorithm is used to control the equivalent system, two weight matrices and are introduced in the Lyapunov function to correspond to the system state control and the disturbance estimation respectively. The LMID algorithm has the effect of disturbance observer, can estimate the equivalent system disturbance, and compensate the matching disturbance of the equivalent system.

[0062] In the embodiment of the application, in S1, the expression of splitting the non-matching disturbance subsystem and the matching disturbance subsystem is:

[0063]

[0064] wherein, represents the derivative of the state of the non-matching disturbance subsystem, represents the derivative of the state of the matching disturbance subsystem, represents the constant system matrix of the non-matching disturbance subsystem, represents the constant system matrix of the matching disturbance subsystem, represents the state variable of the system, represents the constant input matrix of the non-matching disturbance subsystem,​ This represents the steady-state input matrix of the matched perturbation subsystem. Indicates the system's control input, This represents the known terms of the non-matched perturbation subsystem. This represents the known terms of the matched perturbation subsystem. This indicates a non-matching perturbation. This indicates a matching perturbation.

[0065] In this embodiment of the invention, S2 includes the following sub-steps:

[0066] S21. Perform disturbance compensation on the split matching disturbance subsystem;

[0067] S22. Perform disturbance estimation on the matched disturbance subsystem after disturbance compensation;

[0068] S23. Use the control quantity to perform equivalent processing on the matched disturbance subsystem after disturbance compensation and disturbance estimation to obtain the equivalent system.

[0069] In this embodiment of the invention, in S21, the expression for perturbation compensation of the split matched perturbation subsystem is:

[0070] ;

[0071] in, This represents the steady system matrix of the matched perturbation subsystem. Represents the system's state variables. This represents the steady-state input matrix of the matched perturbation subsystem. Indicates reference input. This represents the known terms of the matched perturbation subsystem. This indicates a matching perturbation. This represents the compensation value for matching perturbations. Indicates the system's control input;

[0072] In S22, the discretized expression for perturbation estimation is:

[0073] ;

[0074] in, Indicates the matching perturbation exist The estimated value at time, Indicates to The first step in organizing the quantity. Indicates the matching perturbation The estimated value, Representing discrete time, Indicates to The second process quantity of sorting, a third process quantity a third process quantity a value at a value at a third process quantity a third process quantity a value at a value at

[0075] S23, the expression of the control quantity is:

[0076] ;

[0077] ;

[0078] wherein, denotes a reference control, denotes a pseudo-inverse, denotes a compensation value for a matching disturbance, , is a partial disturbance compensation ratio, denotes a simple expression form of a diagonal matrix.

[0079] In the embodiment of the present application, in S3, the expression of the equivalent system is:

[0080] ;

[0081] wherein, denotes a derivative of a state variable of the control system, denotes a steady system matrix of the control system, denotes a state variable of the control system, denotes a steady input matrix of the control system, denotes a reference control, denotes a known item of the equivalent system, denotes a disturbance of the equivalent system.

[0082] In the embodiment of the present application, in S3, when the original disturbance is bounded and the derivative of the matching disturbance is bounded, then the disturbance of the equivalent system is bounded.

[0083] In the embodiment of the present application, in S41, the expression of the weight matrix of the state control is:

[0084] ;

[0085] wherein, denotes a simple expression form of a diagonal matrix, denotes a state variable weight of the component of the 1st dimension, representing state variables weight of the component of the 1st dimension, representing state variables weight of the component of the 1st dimension, representing state variables weight of the component of the 1st dimension.

[0086] In S41, the expression of the weight matrix of the disturbance estimation is as follows:

[0087] ;

[0088] wherein, representing disturbance weight of the component of the 1st dimension, representing disturbance weight of the component of the 1st dimension, representing state variables representing a simple expression form of a diagonal matrix.

[0089] In the embodiment of the application, the intelligent chassis system comprises a transmission system, a driving system, a steering system, a braking system and a suspension system. For the dynamic control of the intelligent chassis system, different control algorithms are usually used, and the intelligent chassis is realized according to the characteristics of different systems, and the key performance of the vehicle is improved.

[0090] Generally, for the purpose of simplifying the analysis process, the disturbances existing in the system can be integrated into the form of lumped disturbances, and the control system of the chassis dynamics can be expressed in the form of the following state space equation:

[0091] (1);

[0092] In the formula, is the state of the control system; is the input of the control system; is a known term, which can represent a known nonlinearity; is a lumped disturbance, is an external disturbance, and ; , , is a constant matrix.

[0093] In the intelligent chassis system, the lumped disturbance acts on the dynamic system in the form of force or torque, which challenges the robustness of the control algorithm. Generally, the robustness of the system can be improved in two ways.

[0094] One is to use a robust control algorithm to suppress the disturbance. Generally, the following methods are included. H2 / H ∞Control, H2 control and H ∞ Control consists of two algorithms, H2 control aims to reduce the 2-norm of the transfer function from disturbance to controlled output, i.e. to reduce the average response of the controlled output to the disturbance, H ∞ Control aims to reduce the infinity norm of the transfer function, i.e. to reduce the maximum response. Sliding mode control (SMC) reduces the dimension of the system by defining a sliding surface, combines the bound of the actual disturbance, designs a suitable Lyapunov function, and uses a sign function to design a control rate to make the system stable. Adaptive control is generally suitable for slow time-varying disturbances, combines Lyapunov function, and adopts integral method to suppress disturbance, but generally cannot obtain accurate disturbance estimation value. Linear matrix inequality (LMI) based control combines the characteristics of disturbance, and converts the establishment condition of Lyapunov function into a solution problem of linear matrix inequality.

[0095] Secondly, disturbance observer algorithm is used to estimate and compensate the disturbance. Generally, the following methods are included. Extend state observer (ESO) belongs to a link of ADRC control, and the disturbance is extended to the state variable of the system and estimated. A series of disturbance observers based on Lyapunov method, and DOB algorithm, NDOB algorithm, UDE algorithm and GPIO algorithm.

[0096] Robust control algorithm can only resist disturbance to a certain extent, and when the disturbance is too large, H2 / H ∞ Control, LMI and other algorithms may fail to solve or obtain a large feedback gain to make the control amount maximum to cause the system to diverge under inappropriate parameter setting, i.e. only a moderate controller parameter can be selected.

[0097] The disturbance observer algorithm inevitably has an error in the estimation process, which is manifested in that the disturbance estimation usually has a lag, it is difficult to follow high-frequency disturbance, and disturbance compensation is only suitable for matching disturbance. These estimation errors will become new equivalent disturbances of the system and still affect the stability of the system.

[0098] Therefore, the two schemes have certain deficiencies in disturbance suppression.

[0099] In view of the limitations of the above single method, the two methods are used to enhance the robustness of the system, the disturbance observer algorithm is IUDE algorithm, the robust control algorithm is LMID algorithm, and the LMID algorithm additionally has the effect of the disturbance observer, i.e. the whole includes two disturbance observers.

[0100] The IUDE algorithm derived from the classical UDE algorithm is applied to estimate and compensate the matched disturbance firstly, and the discrete expression of the algorithm is given for practical application. In addition, the boundedness of the equivalent system disturbance is analyzed after the IUDE algorithm is applied.

[0101] The LMID algorithm derived from the classical LMI algorithm is applied to control the equivalent system processed by the IUDE algorithm, and the algorithm belongs to the category of robust control algorithm, and can also play the role of disturbance observer to estimate the equivalent system disturbance. The weight is introduced for the state control and disturbance estimation to adjust the flexibility.

[0102] For the control of the intelligent chassis dynamics system, the lumped disturbance can be used to express the unknown system state, sensor error, unmodeled dynamic characteristics or system parameter variation, etc. For example, in the suspension dynamics system, depending on the form of the dynamics modeling, the unknown disturbance can be the vertical velocity of the sprung mass, road excitation or nonlinearity or dead zone of the actuator, etc.

[0103] The algorithm designed in the application is named as LMID-IUDE algorithm, and the controller structure is shown as Figure 2 The following introduces the algorithm from four parts: firstly, the system is defined according to the non-matched / matched disturbance. Then, the disturbance observer algorithm, IUDE algorithm, is introduced to process the matched disturbance. Subsequently, the disturbance form of the equivalent system after the introduction of the IUDE algorithm is analyzed. Finally, the robust control algorithm, LMID algorithm, is introduced to control the equivalent system and estimate the equivalent system disturbance.

[0104] The robustness of the system is enhanced by estimating and compensating the disturbance. However, since the compensation of the disturbance is performed through the control variable , it is only applicable to the compensation of the matched disturbance. Therefore, when the algorithm related to the disturbance observer is introduced, the system needs to be split firstly.

[0105] For the system in formula (1), the non-matched disturbance is defined without loss of generality, wherein is the non-matched disturbance in the channel, is the matched disturbance. The matrix is divided according to the same dimension in the row, , the known term and the lumped disturbance are obtained, , , and , the system (1) is split into the non-matched / matched disturbance subsystem:

[0106] (2);

[0107] In the formula: unmatched / matched disturbances are generally classified according to whether the channel contains control variables, that is, generally there are .

[0108] For the matched perturbation subsystem (second equation) in equation (2), the IUDE algorithm design reference system is as follows:

[0109] (3);

[0110] In the formula: For reference input; This is the compensation value for the matching disturbance; defined. This is an estimate of the matching perturbation; This is a partial disturbance compensation ratio, and .

[0111] definition , , , , , They are respectively , , , , , Laplace transform.

[0112] design ,in And a single diagonal element satisfies , This type of filter It has two adjustable parameters and Among the parameters Parameters that affect the cutoff frequency and amplitude-frequency response of a filter It can directly affect the amplitude-frequency response without affecting the cutoff frequency. Reduce or Both can reduce the amplitude-frequency response of the filter and improve disturbance estimation. The filter parameters are defined as follows: and .

[0113] Applying a Laplace transform to the reference system of equation (2), we obtain:

[0114] (4);

[0115] Define the inverse Laplace transform as ,pass and Two operations are needed to complete the design of the reference system. However, it is difficult to simplify the structure of equation (4) to obtain a feasible disturbance observer expression, so the following equation is introduced for processing:

[0116] (5);

[0117] Definition is the Laplace transform of , and so that is easily simplified and expressed as:

[0118] (6);

[0119] Since and are diagonal matrices, the order of matrix multiplication can be exchanged, so the following relationship is obtained:

[0120] (7);

[0121] That is, can be regarded as the result of filtered by the filter In the controller design, the z-transform can be introduced to obtain the discrete expression, which is substituted into the simplified form to obtain:

[0122] (8);

[0123] Thus, the discrete expression is:

[0124] (9)

[0125] The above analysis is for the reference system. When the matrix is full rank, the pseudo-inverse can be defined, and the reference system can be made equivalent to the actual system by controlling the quantity , and designed as:

[0126] (10);

[0127] Based on the IUDE algorithm, the existing algorithm can be redesigned, which is expressed as applying the existing algorithm to the design of the reference control , estimating and compensating the disturbance by the IUDE algorithm, and obtaining the actual control rate of equation (9).

[0128] For the equivalent system obtained after introducing the IUDE algorithm, the disturbance form of the equivalent system is given below, and the known term Thus, the equivalent system is processed as a linear system form, which is convenient for the design of subsequent algorithm.

[0129] Considering the known term , which is in the matching disturbance channel , can be compensated directly, while , which is in the non-matching disturbance channel, cannot, so is classified as equivalent non-matching disturbance. Define and , and define and , then the equivalent system of the system of equation (1) can be arranged as:

[0130] (11); In the equation:

[0131] , .

[0132] Next, the and are arranged and analyzed for the disturbance of the equivalent system equation (11).

[0133] Since , , , , , are all diagonal matrices, the sub-terms can be studied and analyzed separately, and the scalar sub-terms , , are defined, and the sub-terms of the matching disturbance compensation term and the disturbance estimation term are defined as and . The sub-terms of are arranged, and the following equation can be obtained:

[0134] (12);

[0135] Define and , and the following equation can be obtained:

[0136] (13);

[0137] From equation (6), we can get , so we can get:

[0138] (14)

[0139] ​That is, when and are bounded, since the diagonal elements of the diagonal matrix are positive, it is guaranteed that and are bounded, no matter what takes. Correspondingly, when is bounded, is also bounded.

[0140] Thus, when the original disturbance is bounded and the derivative of the matched disturbance is bounded, the disturbance of the equivalent system represented by equation (11) after the IUDE algorithm is applied is bounded.

[0141] For the equivalent system represented by equation (11) after the IUDE algorithm is applied, for the new equivalent system disturbance , define the disturbance estimate as , and , define the error of the disturbance estimate as . The control rate is designed as:

[0142] (15);

[0143] At this time, , and .

[0144] Define the disturbance estimate as:

[0145] (16);

[0146] From the first equation in equation (16), we have:

[0147] (17);

[0148] Combined with the second equation in equation (16), we have:

[0149] (18);

[0150] And from we get .

[0151] For the system state and the error of the equivalent disturbance estimate , define the weights and , respectively, and define:

[0152] (19);

[0153] Definition , , and define a symmetric positive definite matrix , combining equation (18), we can get:

[0154] (20);

[0155] Since , where the diagonal matrix has positive diagonal elements, equation (20) can be rearranged as:

[0156] (21);

[0157] In equation (21), , denotes the transpose of the previous term, which here denotes .

[0158] Define the scalar , and define , which can be rearranged as:

[0159] (22);

[0160] In equation (22), , , , , denotes .

[0161] By solving equation (22), we get the feedback control rate and , which realizes further estimation of the disturbance.

[0162] In addition, define , , we get the system , and indicates that the system is uniformly bounded.

[0163] This part of the algorithm is named LMID algorithm, and the overall algorithm is named LMID-IUDE algorithm. Since the LMID algorithm also contains a disturbance observer in the design process, the algorithm is also named double disturbance observer algorithm.

[0164] The following will be described in conjunction with specific embodiments.

[0165] In the intelligent chassis system, the 1 / 4 car active suspension control system is selected as the control use case, and its structure is as Figure 2 and Figure 3As shown, it includes two subsystems: the sprung system and the unsprung system, which are related to ride comfort and handling stability, respectively. The dynamic equations are expressed as follows:

[0166] (twenty three);

[0167] In the formula, The vertical displacement of the sprung mass; The vertical displacement of the unsprung mass; For vertical excitation of the road surface; This refers to the dynamic deflection of the suspension. For wheel dynamic deformation; For the sprung mass; Unsprung mass; For spring stiffness; For tire stiffness; For damping of the shock absorber; For tire damping; The actuating force of the actuator, ranging from [-2500, 2500]. Speed Take 20 .

[0168] In the control of the sprung system related to ride comfort, the sprung system equation in equation (23) is usually processed, corresponding to the state variable defined in equation (1) as follows: The control quantity is The disturbance term is And there are: , , , .

[0169] Five comparison algorithms were designed for this system. Algorithm 1 is an uncontrolled algorithm, i.e. Algorithm 2 is based on LMI control. Algorithm 3 uses the IUDE algorithm alone, which only compensates for matched disturbances once and relies on the inherent properties of the system to ensure stability. Algorithm 4 uses the LMID algorithm alone, which considers both matched and unmatched disturbances to solve for feedback gain. It ensures system stability by controlling the pole placement of the control system and estimating and compensating for matched disturbances. Algorithm 5 is the proposed LMID-IUDE algorithm. This algorithm first estimates and compensates for a disturbance once using the IUDE algorithm to obtain the equivalent system, and then processes the equivalent system using the LMID algorithm. It will simultaneously provide the estimation results of the IUDE algorithm for matched disturbances and the estimation results of the LMID algorithm for disturbances in the equivalent system.

[0170] For Algorithm 2, the basic LMI control, its design process is as follows: This system defines a symmetric positive definite matrix. , design constant feedback gain , control rate . design Lyapunov function , define scalar , let , get:

[0171] (24);

[0172] Equation (24) is a nonlinear system about and , in order to solve the above equation, define , , multiply on both sides of the inequality, get:

[0173] (25);

[0174] Thus, it is transformed into a linear matrix inequality about and , combined with the given , solve the equation, get .

[0175] In addition, algorithm 3 and algorithm 5 use the same set of IUDE algorithm parameters, and take Partial disturbance compensation is realized to prevent the control of suspension deflection and wheel dynamic load from being greatly deteriorated, and algorithm 4 and algorithm 5 use the same set of LMID algorithm parameters.

[0176] On the C-level random road, the simulation results of three performance indexes about sprung acceleration , wheel dynamic deformation , suspension deflection are shown in Figure 4 , and the root-mean-square (rms) values are shown in table 1. In algorithm 5, the observation effect of IUDE algorithm on matching disturbance and the observation effect of LMID algorithm on equivalent matching disturbance are shown in Figure 5 .

[0177] Table 1

[0178] Control algorithm Algorithm 1 : Passive suspension 0.459 1.761 5.169 Algorithm 2: Basic LMI control 0.324 1.991 5.075 Algorithm 3: IUDE algorithm 0.377 1.897 5.279 Algorithm 4: LMID algorithm 0.269 2.735 6.935 Algorithm 5: LMID-IUDE algorithm 0.162 4.185 7.646

[0179] Combined with the RMS value and Figure 4 , the sprung acceleration The control effects of the five algorithms from bad to good are algorithm 1, algorithm 3, algorithm 2, algorithm 4 and algorithm 5, which proves the effectiveness of the LMID-IUDE algorithm. The LMID-IUDE algorithm uses the robust control algorithm (LMID algorithm) and the disturbance observer (IUDE algorithm) simultaneously, and the LMID algorithm also has the effect of disturbance estimation, realizes double-disturbance observer control, and has obvious inhibitory effect on disturbance compared with single control method, can effectively ensure the stability of the system, and better improves the system smoothness.

[0180] For the wheel dynamic deformation and the suspension dynamic deflection These two indicators, the designed algorithm exists deterioration, from the smoothness and handling stability control itself exists conflict, also reflects the designed algorithm on the smoothness of the improvement is better.

[0181] From Figure 5 It can be seen that the IUDE algorithm effectively realizes the estimation and compensation of the matching disturbance, but there is still a certain estimation error, which will be converted into the equivalent system matching disturbance, and the disturbance observer designed by the LMID algorithm also better realizes the estimation of the equivalent system disturbance.

[0182] Those skilled in the art will realize that the embodiments described herein are for the purpose of helping the reader to understand the principles of the present application, and should be understood as not limiting the protection scope of the present application to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations according to the technical inspiration disclosed in the present application without departing from the essence of the present application, and these modifications and combinations are still within the protection scope of the present application.

Claims

1. A smart chassis dual-disturbance observer control method, characterized in that, Includes the following steps: S1. The control system is divided into a non-matched disturbance subsystem and a matched disturbance subsystem; S2. The IUDE method is used to perform perturbation compensation and perturbation estimation on the split matched perturbation subsystem to obtain the equivalent system; S3. Determine the boundedness of the disturbance of the equivalent system; S4. Using the weight matrix of state control and the weight matrix of disturbance estimation, perform disturbance estimation and compensation on the equivalent system after determining the boundedness of the disturbance.

2. The intelligent chassis dual-disturbance observer control method according to claim 1, characterized in that, In S1, the expression for separating the unmatched perturbation subsystem and the matched perturbation subsystem is: ； in, Represents the state of the non-matched disturbance subsystem The derivative, Indicates the state of the matched disturbance subsystem The derivative, This represents the steady system matrix of the unmatched perturbation subsystem. This represents the steady system matrix of the matched perturbation subsystem. Represents the system's state variables. This represents the steady-state input matrix of the unmatched perturbation subsystem. This represents the steady-state input matrix of the matched perturbation subsystem. Indicates the system's control input, This represents the known terms of the non-matched perturbation subsystem. This represents the known terms of the matched perturbation subsystem. This indicates a non-matching perturbation. This indicates a matching perturbation.

3. The intelligent chassis dual-disturbance observer control method according to claim 1, characterized in that, S2 includes the following sub-steps: S21. Perform disturbance compensation on the split matching disturbance subsystem; S22. Perform disturbance estimation on the matched disturbance subsystem after disturbance compensation; S23. Use the control quantity to perform equivalent processing on the matched disturbance subsystem after disturbance compensation and disturbance estimation to obtain the equivalent system.

4. The intelligent chassis dual-disturbance observer control method according to claim 1, characterized in that, In step S21, the expression for perturbation compensation of the split matched perturbation subsystem is: ； in, This represents the steady system matrix of the matched perturbation subsystem. Represents the system's state variables. This represents the steady-state input matrix of the matched perturbation subsystem. Indicates reference input. This represents the known terms of the matched perturbation subsystem. This indicates a matching perturbation. This represents the compensation value for matching perturbations. Indicates the system's control input; In step S22, the discretized expression for perturbation estimation is: ； in, Indicates the matching perturbation exist The estimated value at time, Indicates to The first step in organizing the quantity. Indicates the matching perturbation The estimated value, Representing discrete time, Indicates to The second process quantity of sorting, Indicates to The third process quantity of sorting exist The value at time, Indicates to The third process quantity of sorting exist The value at time; In S23, the control quantity The expression is: ； ； in, Indicates reference control, Indicates a false reversal. This represents the compensation value for the matching disturbance.

5. The intelligent chassis dual-disturbance observer control method according to claim 1, characterized in that, In S3, the equivalent system is expressed as follows: ； in, Represents the state variables of the control system The derivative, The matrix representing the steady system of the control system. Represents the state variables of the control system. Represents the steady input matrix of the control system. Indicates reference control, Represents the known terms of the equivalent system. It represents the disturbance of the equivalent system.

6. The intelligent chassis dual-disturbance observer control method according to claim 1, characterized in that, In S3, when the original disturbance is bounded and the derivative of the matched disturbance is bounded, the disturbance of the equivalent system is bounded.

7. The intelligent chassis dual-disturbance observer control method according to claim 1, characterized in that, In S41, the weight matrix for state control The expression is: ； in, A simplified representation of a diagonal matrix. Represents state variables The weights of the components in the first dimension. Represents state variables No. The weights of the components in each dimension. Represents the real number field. Represents state variables The dimension; In S41, the weight matrix for perturbation estimation The expression is: ； in, Indicates disturbance The weights of the components in the first dimension. Indicates disturbance No. The weights of the components in each dimension. A simplified representation of a diagonal matrix.

Images