Compact-form model-free adaptive disturbance compensation control method for immeasurable disturbances

Abstract

The present invention discloses a compact model-free adaptive disturbance compensation control method for immeasurable disturbances. This method establishes a dynamic linearization data model of the controlled object under the action of immeasurable disturbances; constructs and solves a cost function to optimize and update the pseudo-Jacobi input matrix and the pseudo-Jacobi disturbance matrix; designs a compact model-free adaptive disturbance compensation control scheme for immeasurable disturbances; constructs and solves an energy function to optimize and update the compact adaptive input matrix and the compact adaptive disturbance matrix; and uses the control scheme of the present invention to control the controlled object under the action of immeasurable disturbances. The control method of the present invention can significantly weaken the influence of immeasurable disturbances on the actual value of the system output of the controlled object, realize effective tracking of the expected value trajectory, and significantly improve the disturbance compensation control performance.

Description

Technical Field

[0001] The present invention belongs to the field of automatic control, and in particular relates to a compact model-free adaptive disturbance compensation control method for unmeasurable disturbances. Background Art

[0002] Disturbances widely exist in actual control systems, such as most controlled objects in industries such as oil refining, petrochemical, chemical, pharmaceutical, food, papermaking, water treatment, thermal power, metallurgy, cement, rubber, machinery, electrics, transportation, robotics, etc., including reactors, distillation columns, machines, equipment, devices, production lines, workshops, factories, unmanned vehicles, unmanned ships, unmanned aerial vehicles, autonomous mobile robots, etc. In fact, the existence of disturbances usually reduces the control performance of the system, and may cause the instability of the entire system in severe cases, thereby affecting system safety.

[0003] The existing compact model-free adaptive control method was first proposed by Hou Zhongsheng and Jin Shangtai in their co-authored book "Model-Free Adaptive Control - Theory and Applications" (Science Press, 2013, page 92). On this basis, inventions CN107991866A and CN107991865A proposed a decoupling method based on SISO, which solved the control problem of strongly coupled multi-input multi-output systems; inventions CN108132600A and CN108345213A proposed a parameter self-tuning method based on neural networks, which solved the problem of time-consuming and laborious parameter selection; invention CN109782588A proposed a heterofactor control method, which solved the control problem of different control channel characteristics of strongly nonlinear multi-input multi-output systems; invention CN111522235A extended invention CN109782588A and proposed a heterofactor control method with parameter self-tuning, which further solved the problem of time-consuming and laborious tuning of heterofactor parameters. It should be noted that the above-mentioned invention methods have not yet considered the control problem of controlled objects under the action of disturbances.

[0004] For multi-input multi-output controlled objects under the action of unmeasurable disturbances, how to efficiently utilize the input and output data measured in real time by the controlled object, analyze and design a disturbance compensation control method without relying on any mathematical model information, and the designed control method can weaken the influence of unmeasurable disturbances on the actual value of the system output of the controlled object and achieve effective tracking of the expected value of the system output has important industrial application value. To achieve the above goal, the present invention proposes a compact model-free adaptive disturbance compensation control method for unmeasurable disturbances. Summary of the Invention

[0005] To solve the problems existing in the background art, the object of the present invention is to provide a compact model-free adaptive disturbance compensation control method for unmeasurable disturbances. The control method runs on a hardware platform to control a controlled object under the action of unmeasurable disturbances. The controlled object is a multi-input multi-output system including multiple control inputs and multiple system outputs. The control method is characterized by including the following steps:

[0006] Step (1): At the k sampling moment, establish a dynamic linearization data model of the controlled object under the action of unmeasurable disturbances. The dynamic linearization data model of the controlled object includes a pseudo-Jacobi input matrix and a pseudo-Jacobi disturbance matrix

[0007] Step (2): Construct a cost function and use the function extreme value method to solve the cost function, and optimize and update the pseudo-Jacobi input matrix and the pseudo-Jacobi disturbance matrix

[0008] Step (3): Based on the optimized pseudo-Jacobi input matrix and the pseudo-Jacobi disturbance matrix of the dynamic linearization data model of the controlled object, design a compact model-free adaptive disturbance compensation control scheme for unmeasurable disturbances. The control scheme includes a compact adaptive input matrix and a compact adaptive disturbance matrix

[0009] Step (4): Construct an energy function and use the momentum gradient descent method to solve the energy function, and optimize and update the compact adaptive input matrix and the compact adaptive disturbance matrix

[0010] Step (5): Use the control scheme optimized in step (4) for the compact adaptive input matrix and the compact adaptive disturbance matrix to control the controlled object under the action of unmeasurable disturbances, weaken the influence of unmeasurable disturbances on the actual value of the system output of the controlled object, and achieve effective tracking of the expected value of the system output.

[0011] Further, the dynamic linearization data model of the controlled object established at the k sampling moment in step (1) is:

[0012]

[0013] where \(k\) is the sampling time and \(k\) is a positive integer; \(y(k + 1)\) is the actual value vector of the system output of the controlled object at the \((k + 1)\)-th sampling time, \(y(k + 1)=[y_1(k + 1),\cdots,y n (k + 1)] T , \(\Delta y(k + 1)=y(k + 1)-y(k)\); \(n\) is the total number of system outputs of the controlled object and \(n\) is an integer greater than 1; \(u(k)\) is the control input vector of the controlled object at the \(k\)-th sampling time, \(u(k)=[u_1(k),\cdots,u m (k)] T , \(\Delta u(k)=u(k)-u(k - 1)\); \(m\) is the total number of control inputs of the controlled object and \(m\) is an integer greater than 1; 1 q×1 = [1; 1; \(\cdots\); 1] q×1 , \(q\) is the total number of unmeasurable disturbances received by the controlled object and \(q\) is a positive integer; is the pseudo-Jacobi input matrix at the \(k\)-th sampling time, is the pseudo-Jacobi disturbance matrix at the \(k\)-th sampling time.

[0014] In step (2), constructing the cost function and using the function extremum method to solve the cost function to optimize and update the pseudo-Jacobi input matrix and the pseudo-Jacobi disturbance matrix mainly includes the following steps:

[0015] Step (2.1): Construct a cost function for the pseudo-Jacobi input matrix

[0016]

[0017] where \(\mu_1\) is the first weight factor;

[0018] Step (2.2): Construct a cost function for the pseudo-Jacobi disturbance matrix

[0019]

[0020] where \(\mu_2\) is the second weight factor;

[0021] Step (2.3): Use the function extremum method to solve the cost function in step (2.1) to optimize and update the pseudo-Jacobi input matrix

[0022]

[0023] where \(\alpha_1\) is the first step factor;

[0024] Step (2.4): Solve the cost function described in step (2.2) using the function extreme value method, and optimize and update the pseudo-Jacobian perturbation matrix

[0025]

[0026] where α2 is the second step size factor.

[0027] Based on the optimization of the pseudo-Jacobian input matrix described in step (2), and the pseudo-Jacobian perturbation matrix the dynamic linearization data model of the controlled object after that, design a compact-form model-free adaptive disturbance compensation control scheme for the unmeasurable disturbance as follows:

[0028]

[0029] where e(k) is the system error vector of the controlled object at the kth sampling moment, e(k)=y * (k)-y(k), e(k)=[e1(k),…,e n (k)] T , Δe(k)=e(k)-e(k - 1); is the compact-form adaptive input matrix at the kth sampling moment, is the compact-form adaptive disturbance matrix at the kth sampling moment.

[0030] Construct the energy function described in step (4) and use the momentum gradient descent method to solve the energy function, optimize and update the compact-form adaptive input matrix described in step (3) and the compact-form adaptive disturbance matrix mainly includes the following steps:

[0031] Step (4.1): Construct the energy function

[0032]

[0033] where y * (k + 1) is the system output expected value vector of the controlled object at the (k + 1)th sampling moment, λ is the penalty factor;

[0034] Step (4.2): Use the momentum gradient descent method to solve the energy function described in step (4.1), optimize and update the compact-form adaptive input matrix

[0035]

[0036] where σ1 is the first learning rate and η1 is the first momentum factor; The partial derivative of the energy function W with respect to ;

[0037] Step (4.3): Use the momentum gradient descent method to solve the energy function described in step (4.1), and optimize and update the tight-format adaptive perturbation matrix

[0038]

[0039] where σ2 is the second learning rate and η2 is the second momentum factor; The partial derivative of the energy function W with respect to ;

[0040] The partial derivative calculation formula of the energy function W with respect to described in step (4.2) is:

[0041]

[0042] The partial derivative calculation formula of the energy function W with respect to described in step (4.3) is:

[0043]

[0044] The mathematical calculation formula is:

[0045] Using the control scheme after optimizing the tight-format adaptive input matrix and the tight-format adaptive perturbation matrix described in step (5) to control the controlled object under the action of the unmeasurable disturbance, including the following steps at each sampling time k:

[0046] Step (5.1): Obtain the system output expected value vector y * (k), the system output actual value vector y(k), and calculate the system error vector e(k) at the current sampling time;

[0047] Step (5.2): Based on step (5.1), use the control scheme after optimizing the tight-format adaptive input matrix and the tight-format adaptive perturbation matrix to calculate the control input vector u(k) at the current sampling time;

[0048] Step (5.3): After the control input vector acts on the controlled object, the actual value vector of the system output of the controlled object at the next sampling moment is obtained.

[0049] Furthermore, the present invention adopts the following technical solutions:

[0050] A non-transitory computer-readable storage medium stores a computer program thereon, characterized in that when the computer program is executed by a processor, the above-mentioned compact-form model-free adaptive disturbance compensation control method for unmeasurable disturbances is implemented.

[0051] Even further, the present invention adopts the following technical solutions:

[0052] An electronic device includes a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that when the processor executes the program, the above-mentioned compact-form model-free adaptive disturbance compensation control method for unmeasurable disturbances is implemented.

[0053] Based on the theoretical basis of the existing compact-form model-free adaptive control method, some inventive methods have made progress in solving problems such as strong coupling of the controlled object, different channel characteristics, and time-consuming and laborious parameter tuning. However, these inventive methods have not considered the control problem of the controlled object under the action of disturbances, restricting their popularization and application. For a multi-input multi-output controlled object under the action of unmeasurable disturbances, the present invention can efficiently utilize the input and output data measured in real time by the controlled object, analyze and design a disturbance compensation control method without relying on any mathematical model information, and the designed control method can weaken the influence of unmeasurable disturbances on the actual value of the system output of the controlled object, realize effective tracking of the expected value of the system output, and has important industrial application value. Description of the Drawings

[0054] Figure 1 is the algorithm principle block diagram of the present invention;

[0055] Figure 2 is the engineering application system block diagram of the present invention;

[0056] Figure 3 is the schematic diagram of the hardware platform for running the present invention;

[0057] Figure 4 is the control effect diagram of the first system output when the compact-form model-free adaptive disturbance compensation control method for unmeasurable disturbances proposed by the present invention and a comparative control method are adopted for a two-input two-output system;

[0058] Figure 5The control effect diagram of the second system output when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a two-input two-output system;

[0059] Figure 6 The first control input curve when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a two-input two-output system;

[0060] Figure 7 The second control input curve when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a two-input two-output system;

[0061] Figure 8 The freezing cycle flow chart of a vapor compression refrigeration system;

[0062] Figure 9 The graphs of two unmeasurable disturbances received by a vapor compression refrigeration system;

[0063] Figure 10 The control effect diagram of the first system output when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a vapor compression refrigeration system;

[0064] Figure 11 The control effect diagram of the second system output when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a vapor compression refrigeration system;

[0065] Figure 12 The first control input curve when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a vapor compression refrigeration system;

[0066] Figure 13 The second control input curve when the compact model-free adaptive disturbance compensation control method and the comparative control method for unmeasurable disturbances proposed by the present invention are adopted in a vapor compression refrigeration system. Specific embodiments

[0067] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0068] Figure 1The algorithm principle block diagram of the present invention is given. The present invention discloses a compact model-free adaptive disturbance compensation control method for unmeasurable disturbances. This method establishes a dynamic linearization data model of the controlled object under the action of unmeasurable disturbances, and this model includes a pseudo-Jacobi input matrix and a pseudo-Jacobi disturbance matrix; constructs and solves a cost function to optimize and update the pseudo-Jacobi input matrix and the pseudo-Jacobi disturbance matrix; designs a compact model-free adaptive disturbance compensation control scheme for unmeasurable disturbances, and this scheme includes a compact adaptive input matrix and a compact adaptive disturbance matrix; constructs and solves an energy function to optimize and update the compact adaptive input matrix and the compact adaptive disturbance matrix; uses the control scheme of the present invention to control the controlled object under the action of unmeasurable disturbances. Next, the implementation steps of the compact model-free adaptive disturbance compensation control method for unmeasurable disturbances provided by the present invention are further described as follows:

[0069] The control method runs on a hardware platform to control the controlled object under the action of unmeasurable disturbances. The controlled object is a multi-input multi-output system including multiple control inputs and multiple system outputs. The control method is characterized by including the following steps:

[0070] Step (1): At the k sampling moment, establish a dynamic linearization data model of the controlled object under the action of unmeasurable disturbances. The dynamic linearization data model of the controlled object includes a pseudo-Jacobi input matrix and a pseudo-Jacobi disturbance matrix

[0071] Step (2): Construct a cost function and use the function extreme value method to solve the cost function to optimize and update the pseudo-Jacobi input matrix in step (1) and the pseudo-Jacobi disturbance matrix

[0072] Step (3): Based on the dynamic linearization data model of the controlled object after optimizing the pseudo-Jacobi input matrix and the pseudo-Jacobi disturbance matrix in step (2), design a compact model-free adaptive disturbance compensation control scheme for unmeasurable disturbances. The control scheme includes a compact adaptive input matrix and a compact adaptive disturbance matrix and a compact adaptive disturbance matrix

[0073] Step (4): Construct an energy function and use the momentum gradient descent method to solve the energy function to optimize and update the compact adaptive input matrix in step (3) and the compact adaptive disturbance matrix

[0074] Step (5): Optimize the tight - format adaptive input matrix using Step (4). and the tight - format adaptive perturbation matrix to control the controlled object under the action of the unmeasurable perturbation, weaken the influence of the unmeasurable perturbation on the actual value of the system output of the controlled object, and achieve effective tracking of the expected value of the system output.

[0075] Furthermore, in Step (1), the dynamic linearization data model of the controlled object under the action of the unmeasurable perturbation established at the k - th sampling moment is as follows:

[0076]

[0077] where k is the sampling moment and k is a positive integer; y(k + 1) is the vector of the actual value of the system output of the controlled object at the (k + 1)-th sampling moment, y(k + 1)=[y1(k + 1),…,y n (k + 1)] T , Δy(k + 1)=y(k + 1)-y(k); n is the total number of system outputs of the controlled object and n is an integer greater than 1; u(k) is the control input vector of the controlled object at the k - th sampling moment, u(k)=[u1(k),…,u m (k)] T , Δu(k)=u(k)-u(k - 1); m is the total number of control inputs of the controlled object and m is an integer greater than 1; 1 q×1 =[1; 1;…; 1] q×1 , q is the total number of unmeasurable perturbations received by the controlled object and q is a positive integer; is the pseudo - Jacobian input matrix at the k - th sampling moment, is the pseudo - Jacobian perturbation matrix at the k - th sampling moment.

[0078] The construction of the cost function in Step (2) and the solution of the cost function using the function extremum method to optimize and update the pseudo - Jacobian input matrix and the pseudo - Jacobian perturbation matrix mainly includes the following steps:

[0079] Step (2.1): Construct a cost function for the pseudo - Jacobian input matrix

[0080]

[0081] where μ1 is the first weight factor;

[0082] Step (2.2): Construct a cost function for the pseudo - Jacobian perturbation matrix

[0083]

[0084] Among them, μ2 is the second weight factor;

[0085] Step (2.3): Solve the cost function described in step (2.1) using the function extremum method, and optimize and update the pseudo-Jacobian input matrix

[0086]

[0087] Among them, α1 is the first step size factor;

[0088] Step (2.4): Solve the cost function described in step (2.2) using the function extremum method, and optimize and update the pseudo-Jacobian perturbation matrix

[0089]

[0090] Among them, α2 is the second step size factor.

[0091] The pseudo-Jacobian input matrix optimized based on step (2) described in step (3) and the pseudo-Jacobian perturbation matrix The dynamic linearization data model of the controlled object after that, the design of a compact model-free adaptive disturbance compensation control scheme for the unmeasured disturbance is:

[0092]

[0093] Among them, e(k) is the system error vector of the controlled object at the kth sampling moment, e(k)=y * (k)-y(k), e(k)=[e1(k),…,e n (k)] T , Δe(k)=e(k)-e(k - 1); is the compact adaptive input matrix at the kth sampling moment, is the compact adaptive disturbance matrix at the kth sampling moment.

[0094] The construction of the energy function described in step (4) and the use of the momentum gradient descent method to solve the energy function, optimize and update the compact adaptive input matrix described in step (3) and the compact adaptive disturbance matrix mainly includes the following steps:

[0095] Step (4.1): Construct the energy function

[0096]

[0097] where y * (k + 1) is the expected value vector of the system output of the controlled object at the (k + 1)-th sampling instant, λ is the penalty factor;

[0098] Step (4.2): Solve the energy function described in step (4.1) using the momentum gradient descent method to optimize and update the tight-format adaptive input matrix

[0099]

[0100] where σ1 is the first learning rate and η1 is the first momentum factor; is the partial derivative of the energy function W with respect to ;

[0101] Step (4.3): Solve the energy function described in step (4.1) using the momentum gradient descent method to optimize and update the tight-format adaptive perturbation matrix

[0102]

[0103] where σ2 is the second learning rate and η2 is the second momentum factor; is the partial derivative of the energy function W with respect to ;

[0104] The partial derivative calculation formula of the energy function W with respect to in step (4.2) is:

[0105]

[0106] The partial derivative calculation formula of the energy function W with respect to in step (4.3) is:

[0107]

[0108] The mathematical calculation formula is:

[0109] The control scheme for the controlled object under the action of the unmeasured disturbance after optimizing the tight-format adaptive input matrix and the tight-format adaptive perturbation matrix in step (5) includes the following steps at each sampling instant k:

[0110] Step (5.1): Obtain the expected value vector y * (k) of the system output at the current sampling moment and the actual value vector y(k) of the system output, and calculate the system error vector e(k) at the current sampling moment;

[0111] Step (5.2): Based on Step (5.1), optimize the tight-form adaptive input matrix by using Step (4) The tight-form adaptive disturbance matrix and calculate the control input vector u(k) at the current sampling moment according to the control scheme after optimization;

[0112] Step (5.3): After the control input vector acts on the controlled object, obtain the actual value vector of the system output of the controlled object at the next sampling moment.

[0113] Figure 2 This is the block diagram of the engineering application system of the present invention. Further, for Figure 2 the hardware platform in the block diagram of the engineering application system, Figure 3 the schematic diagram of the hardware platform for running the present invention is given. Specifically, the present invention uses a non-transitory computer-readable storage medium, on which a computer program is stored. The computer program, when executed by a processor, implements the above-mentioned tight-form model-free adaptive disturbance compensation control method for unmeasurable disturbances; the present invention uses an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. The processor, when executing the program, implements the above-mentioned tight-form model-free adaptive disturbance compensation control method for unmeasurable disturbances.

[0114] The following are two specific embodiments of the present invention. Specific Embodiment 1:

[0116] The controlled object adopts a two-input two-output nonlinear system:

[0117]

[0118]

[0119]

[0120]

[0121] y1(k + 1) = x 11 (k + 1)

[0122] y2(k + 1) = x 21 (k + 1)

[0123] Among them, a(k) = 1 + 0.1sin(2πk / 1500) and b(k) = 1 + 0.1cos(2πk / 1500) are two time-varying parameters;

[0124]

[0125] d1(k) and d2(k) are immeasurable disturbances. It should be noted that the given immeasurable disturbances are used to generate input-output data for the two-input two-output nonlinear system, and they are not used for controller design itself. Therefore, the controlled two-input two-output nonlinear system is a two-input two-output nonlinear system under the action of immeasurable disturbances.

[0126] The expected value trajectory y of the system output * (k) is as follows:

[0127]

[0128]

[0129] In the specific embodiment 1, m = n = q = 2.

[0130] To more clearly compare the control performance of the control method of the present invention and the comparative control method, the Integral Time-weighted Absolute Error (ITAE) is used as the control performance evaluation index:

[0131]

[0132] Among them, is the expected value of the j-th system output at the k-th sampling moment, and y j (k) is the actual value of the j-th system output at the k-th sampling moment, j = 1,..., n. The smaller the value of ITAE(e j ), it indicates that the overall error between the actual value y j (k) of the j-th system output and the expected value of the j-th system output is smaller, the control accuracy and speed are higher, and the control performance is better.

[0133] The hardware platform for running the control method of the present invention uses an industrial control computer.

[0134] The control method of the present invention is used to control the two-input two-output system. The parameter settings of the control method of the present invention are: 0.15, -0.3], α1 = 0.3, α2 = 0.2, μ1 = 3, μ2 = 1, σ1 = 0.1, σ2 = 0.9, η1 = 0.6, η2 = 0.1, λ = 0.4.

[0135] When using the control method of the present invention to control a two-input two-output system under an unmeasurable disturbance, the following steps are included at each sampling time k: a) Obtain the expected value vector y * (k) of the system output and the actual value vector y(k) of the system output at the current sampling time, and calculate the system error vector e(k) at the current sampling time; b) Based on step a), use step (4) to optimize the tight-form adaptive input matrix and the tight-form adaptive disturbance matrix and calculate the control input vector u(k) at the current sampling time according to the control scheme after optimization; c) After the control input vector acts on the two-input two-output system, obtain the actual value vector of the system output of the two-input two-output system at the next sampling time; d) Repeat a) to c) until the sampling time ends.

[0136] The comparison of the control effects between the control method of the present invention and the existing PID control method (comparative control method) is as follows: Figure 4 It is the control effect diagram of the first system output when using the control method of the present invention and the comparative control method, Figure 5 It is the control effect diagram of the second system output when using the control method of the present invention and the comparative control method, Figure 6 It is the first control input curve when using the control method of the present invention and the comparative control method, Figure 7 It is the second control input curve when using the control method of the present invention and the comparative control method; From the perspective of control performance evaluation indicators, the ITAE (e1) of the first system output using the control method of the present invention is 23511, and the ITAE (e2) of the second system output is 9732. The ITAE (e1) of the first system output using the comparative control method is 32190, and the ITAE (e2) of the second system output is 17413. The results of the control performance evaluation indicators are listed in Table 1; From the perspective of the system output curve, the control method of the present invention can effectively suppress the influence of the unmeasurable disturbance on the actual value of the system output of the two-input two-output system, and the control performance of the control method of the present invention is better than that of the comparative control method. Based on the above investigation, it fully shows that the tight-form model-free adaptive disturbance compensation control method for unmeasurable disturbances provided by the present invention can significantly weaken the influence of unmeasurable disturbances on the actual value of the system output of the controlled object, realize effective tracking of the expected value trajectory, and significantly improve the disturbance compensation control performance.

[0137] Table 1 Comparison of control performance of two-input two-output system

[0138] Specific embodiment 2:

[0140] Vapor Compression Refrigeration Systems (VCRS) are the most common refrigeration cycle devices, widely used in households (such as household refrigerators and air conditioners), commercial (such as building and automotive air conditioners, cold storage warehouses), and industrial (such as petrochemical plants, natural gas processing plants). The refrigeration cycle process is as shown in Figure 8 shown. The two disturbances in the refrigeration cycle process are the inlet temperature of the cooling medium and the inlet temperature of the medium to be cooled. Today, with the extensive use of high-energy-consuming refrigeration equipment, achieving disturbance compensation control for vapor compression refrigeration systems is of great significance for promoting energy conservation and consumption reduction in China and even the world.

[0141] The controlled object, the vapor compression refrigeration system, is a two-input two-output nonlinear system. The two control inputs u1 and u2 of the vapor compression refrigeration system are the compressor frequency (Hz) and the valve opening (%) respectively. The two system outputs y1 and y2 of the vapor compression refrigeration system are the superheat degree (°C) and the outlet temperature of the medium to be cooled (°C) respectively. The two disturbances d1 and d2 received by the vapor compression refrigeration system are the inlet temperature of the cooling medium (°C) and the inlet temperature of the medium to be cooled (°C). d1 and d2 are not measured online using corresponding temperature sensors and are unmeasurable disturbances. Figure 9 It is a graph of the two unmeasurable disturbances received by the vapor compression refrigeration system. It should be noted that the given unmeasurable disturbances are for generating input-output data for the vapor compression refrigeration system and are not used for controller design itself. Therefore, the controlled object, the vapor compression refrigeration system, is a two-input two-output nonlinear system under the action of unmeasurable disturbances. In the specific embodiment 2, m = n = q = 2. The hardware platform for running the control method of the present invention uses an industrial control computer.

[0142] The initial working conditions of the vapor compression refrigeration system as the controlled object are: u1(0) = 36.45 Hz, u2(0) = 48.79%, y1(0) = 14.65 °C, y2(0) = -22.15 °C. To meet the refrigeration demand of the medium to be cooled, the expected value trajectory y1* of the system output is step-adjusted from 14.65 °C to 7.2 °C at the 2nd minute, from 7.2 °C to 22.2 °C at the 9th minute, and finally from 22.2 °C to 11.65 °C at the 16th minute. The expected value trajectory of the system output is step-adjusted from -22.15 °C to -22.65 °C at the 2nd minute.

[0143] The control method of the present invention is used to control the vapor compression refrigeration system. The parameter settings of the control method of the present invention are: α1 = 0.5, α2 = 0.5, μ1 = 1, μ2 = 1, σ1 = 0.1, σ2 = 0.9, η1 = 0.2, η2 = 0.1, λ = 0.1.

[0144] When the control method of the present invention is used to control the vapor compression refrigeration system under the action of an unmeasurable disturbance, at each sampling moment k, the following steps are included: a) Obtain the expected value vector y * (k) of the system output and the actual value vector y(k) of the system output at the current sampling moment, and calculate the system error vector e(k) at the current sampling moment; b) Based on step a), use step (4) to optimize the tight-form adaptive input matrix and the tight-form adaptive disturbance matrix to calculate the control input vector u(k) at the current sampling moment according to the control scheme after that; c) After the control input vector acts on the vapor compression refrigeration system, obtain the actual value vector of the system output of the vapor compression refrigeration system at the next sampling moment; d) Repeat a) to c) until the sampling moment ends.

[0145] The comparison of the control effects of the control method of the present invention and the existing PID control method (comparative control method) is as follows: Figure 10 It is the control effect diagram of the first system output when the control method of the present invention and the comparative control method are used, Figure 11 It is the control effect diagram of the second system output when the control method of the present invention and the comparative control method are used, Figure 12 It is the first control input curve when the control method of the present invention and the comparative control method are used, Figure 13 It is the second control input curve when the control method of the present invention and the comparative control method are used; from the perspective of the control performance evaluation index, the ITAE(e1) of the first system output using the control method of the present invention is 291180, and the ITAE(e2) of the second system output is 5732. The ITAE(e1) of the first system output using the comparative control method is 506970, and the ITAE(e2) of the second system output is 52244. The results of the control performance evaluation index are listed in Table 2; from the perspective of the system output curve, the control method of the present invention can effectively suppress the influence of the unmeasurable disturbance on the actual value of the system output of the vapor compression refrigeration system, and the control performance of the control method of the present invention is better than that of the comparative control method. Based on the above investigations, it fully shows that the tight-form model-free adaptive disturbance compensation control method for unmeasurable disturbances provided by the present invention can significantly weaken the influence of unmeasurable disturbances on the actual value of the system output of the controlled object, realize the effective tracking of the expected value trajectory, and significantly improve the disturbance compensation control performance.

[0146] Table 2 Comparison of the control performance of the vapor compression refrigeration system

[0147]

[0148] Furthermore, the following two points should be specifically pointed out:

[0149] (1) Disturbances widely exist in actual control systems, such as most controlled objects in industries such as oil refining, petrochemical, chemical, pharmaceutical, food, paper making, water treatment, thermal power, metallurgy, cement, rubber, machinery, electrical, transportation, robotics, etc., including reactors, distillation columns, machines, equipment, devices, production lines, workshops, factories, unmanned vehicles, unmanned ships, unmanned aerial vehicles, autonomous mobile robots, etc. For example, the vapor compression refrigeration system will be continuously and complexly affected by two unmeasurable disturbances, namely the inlet temperature of the cooling medium and the inlet temperature of the medium to be cooled. Specific embodiment 2 shows that the control method of the present invention can significantly weaken the influence of unmeasurable disturbances on the actual output value of the controlled object system, realize effective tracking of the expected value trajectory, and thus significantly improve the disturbance compensation control performance. Another example is that an unmanned ship is extremely vulnerable to the influence of the water surface wind field during operation. Changes in wind speed and direction will not only affect the speed and heading of the unmanned ship, but may even cause the unmanned ship to capsize in severe cases; when the water surface wind field shows turbulent characteristics due to complex environmental influences, the wind speed and direction show a random and irregular motion form, becoming unmeasurable disturbances. Using the control method of the present invention can achieve compensation for this unmeasurable disturbance and realize the stable operation of the unmanned ship, which is of great significance for improving the safety and reliability of the unmanned ship.

[0150] (2) In the above specific embodiment 1 and specific embodiment 2, the hardware platform for running the control method of the present invention is an industrial control computer; in actual applications, according to specific circumstances, any one or any combination of single-chip microcomputer controllers, microprocessor controllers, field programmable gate array controllers, digital signal processing controllers, embedded system controllers, programmable logic controllers, distributed control systems, fieldbus control systems, industrial Internet of Things control systems, and industrial Internet control systems can also be selected as the hardware platform for running the control method of the present invention.

[0151] Through the description of the above embodiments, those skilled in the art can clearly understand that the implementation of the present invention can be achieved by means of software plus a necessary hardware platform. Embodiments of the present invention can be implemented using existing processors, or by dedicated processors used for this purpose or other purposes in a suitable system, or by a hardwired system. Embodiments of the present invention also include non-transitory computer-readable storage media, which include machine-readable media for carrying or having machine-executable instructions or data structures stored thereon; such machine-readable media can be any available media accessible by a general or special-purpose computer or other machine with a processor. For example, such machine-readable media can include RAM, ROM, EPROM, EEPROM, CD-ROM or other optical disk memories, disk memories or other magnetic storage devices, or any other medium that can be used to carry or store the required program code in the form of machine-executable instructions or data structures and can be accessed by a general or special-purpose computer or other machine with a processor. When information is transmitted or provided to a machine through a network or other communication connection (hardwired, wireless, or a combination of hardwired and wireless), this connection is also regarded as a machine-readable medium.

[0152] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the protection scope of the present invention.

Claims

1. Compact-form model-free adaptive disturbance compensation control method for unmeasurable disturbances. The control method runs on a hardware platform to control a controlled object under the action of unmeasurable disturbances. The controlled object is a multi-input multi-output system with multiple control inputs and multiple system outputs. The control method is characterized by the following steps: Step (1): At the k sampling instant, establish a dynamic linearization data model of the controlled object under the action of an unmeasurable disturbance, where the dynamic linearization data model of the controlled object includes a pseudo-Jacobi input matrix and a pseudo-Jacobi disturbance matrix Step (2): Construct a cost function and use the function extreme value method to solve the cost function, and optimize and update the pseudo-Jacobi input matrix in step (1) and the pseudo-Jacobi perturbation matrix Step (3): Optimize the pseudo-Jacobian input matrix based on Step (2). and the pseudo-Jacobian perturbation matrix After the dynamic linearization data model of the controlled object, design a compact model-free adaptive perturbation compensation control scheme for the unmeasurable perturbation as follows: where k is the sampling time and k is a positive integer; u(k) is the control input vector of the controlled object at the k-th sampling time, u(k)=[u1(k),…,u m (k)] T ; e(k) is the system error vector of the controlled object at the k-th sampling time, e(k)=y * (k)-y(k), e(k)=[e1(k),…,e n (k)] T , Δe(k)=e(k)-e(k - 1); y * (k) is the expected value vector of the system output of the controlled object at the k-th sampling time, y(k) is the actual value vector of the system output of the controlled object at the k-th sampling time, y(k)=[y1(k),…,y n (k)] T ; is the compact-form adaptive input matrix at the k-th sampling time; is the compact-form adaptive disturbance matrix at the k-th sampling time; 1 q×1 =[1; 1;…; 1] q×1 , q is the total number of unmeasurable disturbances received by the controlled object, q is a positive integer; m is the total number of control inputs of the controlled object, m is an integer greater than 1; n is the total number of system outputs of the controlled object, n is an integer greater than 1; Step (4): Construct an energy function and use the momentum gradient descent method to solve the energy function, and optimize and update the tight-format adaptive input matrix in step (3) and the tight-format adaptive perturbation matrix Step (5): Optimize the tight format adaptive input matrix using Step (4). and the tight format adaptive disturbance matrix Use the control scheme after that to control the controlled object under the action of the unmeasurable disturbance, weaken the influence of the unmeasurable disturbance on the actual value of the system output of the controlled object, and achieve effective tracking of the expected value of the system output.

2. The compact model-free adaptive disturbance compensation control method for immeasurable disturbances according to claim 1, wherein In step (1), at the k-th sampling moment, a dynamic linearization data model of the controlled object under the action of unmeasurable disturbances is established as: where, Δy(k + 1) = y(k + 1) - y(k); Δu(k) = u(k) - u(k - 1); is the pseudo-Jacobian input matrix at the k-th sampling instant; is the pseudo-Jacobian perturbation matrix at the k-th sampling instant.

3. The compact model-free adaptive disturbance compensation control method for immeasurable disturbances according to claim 2, wherein Constructing the cost function described in step (2) and solving the cost function by using the function extreme value method to optimize and update the pseudo-Jacobi input matrix described in step (1) and the pseudo-Jacobi perturbation matrix mainly includes the following steps: Step (2.1): For the pseudo-Jacobian input matrix Construct a cost function where μ1 is the first weight factor; Step (2.2): For the pseudo-Jacobian perturbation matrix Construct a cost function where μ2 is the second weight factor; Step (2.3): Solve the cost function described in step (2.1) using the function extreme value method, and optimize and update the pseudo-Jacobi input matrix where α1 is the first step factor; Step (2.4): Solve the cost function described in step (2.2) using the function extreme value method to optimize and update the pseudo-Jacobian perturbation matrix where α2 is the second step factor.

4. The compact model-free adaptive disturbance compensation control method for immeasurable disturbances according to claim 1, characterized in that Constructing an energy function and solving the energy function using the momentum gradient descent method to optimize and update the tight-form adaptive input matrix described in step (3) and the tight-form adaptive perturbation matrix mainly includes the following steps: Step (4.1): Construct an energy function where λ is a penalty factor; Step (4.2): Solve the energy function described in step (4.1) using the momentum gradient descent method to optimize and update the tight format adaptive input matrix where σ1 is the first learning rate and η1 is the first momentum factor; is the partial derivative of the energy function W with respect to ; Step (4.3): Solve the energy function described in step (4.1) using the momentum gradient descent method to optimize and update the tight-form adaptive perturbation matrix wherein, σ2 is the second learning rate, and η2 is the second momentum factor; is the partial derivative of the energy function W with respect to partial derivative.

5. The compact model-free adaptive disturbance compensation control method for immeasurable disturbances according to claim 4, characterized in that The partial derivative calculation formula of the energy function W described in step (4.2) with respect to is as follows: The partial derivative calculation formula of the energy function W described in step (4.3) with respect to is as follows:

6. The compact model-free adaptive disturbance compensation control method for immeasurable disturbances according to claim 5, characterized in that, The mathematical calculation formula is as follows:

7. The compact model-free adaptive disturbance compensation control method for immeasurable disturbances according to claim 1, wherein Optimizing the tight-format adaptive input matrix by using step (4) as described in step (5) and the tight-format adaptive perturbation matrix The controlled object is controlled under the action of an immeasurable disturbance by the control scheme after that, and at each sampling instant k, the following steps are included: Step (5.1): Obtain the expected value vector y * (k) of the system output at the current sampling moment and the actual value vector y(k) of the system output, and calculate the system error vector e(k) at the current sampling moment; Step (5.2): Based on step (5.1), use step (4) to optimize the tight-format adaptive input matrix and the tight-format adaptive perturbation matrix to calculate the control input vector u(k) at the current sampling time according to the control scheme after optimization; Step (5.3): After the control input vector acts on the controlled object, the actual value vector of the system output of the controlled object at the next sampling moment is obtained.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the compact-form model-free adaptive disturbance compensation control method for unmeasurable disturbances according to any one of claims 1 to 7.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the compact-form model-free adaptive disturbance compensation control method for unmeasurable disturbances according to any one of claims 1 to 7.

Images