Method for optimizing gain of LADRC controller

Abstract

The invention discloses a method for optimizing the gain of an LADRC controller. The method comprises the following steps: S1, defining a gain adjustment function of the controller; s2, setting and optimizing a gain adjustment function; in the step S1, a proportional gain and a differential gain are defined. In the step S1, firstly, the phase error at the current moment, the change rate of the phase error and the change acceleration of the phase error are defined, and then the proportional gain and the differential gain are defined respectively. The invention aims to provide a method for optimizing the gain of an LADRC controller, and aims to improve the control performance of the LADRC controller in a complex system environment, especially to enhance the robustness, self-adaptability and accurate tracking capability of the LADRC controller in dynamic disturbance, system nonlinearity and complex environments.

Description

Technical Field

[0001] The present invention relates to the fields of power system control technology and automation control technology, and in particular to a method for optimizing the gain of a LADRC controller that can be applied to scenarios such as microgrids, distributed power generation, and rural power grids. Background Art

[0002] Linear Active Disturbance Rejection Control (LADRC) controller has been widely used in various complex power systems due to its strong nonlinear suppression ability and good robustness, especially in the field of microgrid and distributed generation. The advantage of LADRC controller is that it can suppress disturbances and nonlinear effects in the system at the same time, and does not require an accurate system model, so it has strong adaptability in practical applications.

[0003] In practical applications, the performance of the LADRC controller is closely related to its gain parameters. If the gain parameters are not selected properly, it may lead to unsatisfactory control effects, such as slow system response or excessive system oscillation. Therefore, how to optimize the gain of the LADRC controller for different power system scenarios to improve the control effect is an important direction of current research. From literature such as "Research on Control Methods Based on Fuzzy PID Controller", it can be seen that this type of existing technology is difficult to cope with dynamic disturbances, system nonlinearity, and accurate tracking of phase errors in complex environments. Therefore, how to design a more efficient and accurate gain optimization method that can not only improve computational efficiency but also avoid local optimal problems is still the key to improving the performance of the LADRC controller.

[0004] In order to optimize the gain of the LADRC controller, fuzzy control is combined with LADRC to dynamically optimize the gain of the LADRC controller, aiming to improve the control performance of the LADRC controller in a complex system environment. This method enhances the adaptability, anti-disturbance and robustness of the LADRC controller through fuzzy control, especially in dynamic disturbances, system nonlinearity and complex environments, and can improve the precise tracking capability. This makes this method suitable for application scenarios such as power systems, microgrids and distributed generation. Summary of the invention

[0005] The purpose of the present invention is to provide a method for optimizing the gain of a LADRC controller, aiming to improve the control performance of the LADRC controller in a complex system environment, especially to enhance its robustness, adaptability and precise tracking capabilities under dynamic disturbances, system nonlinearity, and complex environments.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A method for optimizing the gain of a LADRC controller comprises the following steps:

[0008] Step S1: defining a gain adjustment function of a controller;

[0009] Step S2: setting and optimizing the gain adjustment function;

[0010] In step S1, define the proportional gain k p (t) and the differential gain k d (t).

[0011] In step S1, first define the phase error at the current moment Phase error rate of change And the phase error change acceleration Then define the proportional gain k p (t) and the differential gain k d (t) is:

[0012]

[0013] in, and It is a gain adjustment function that is adjusted in real time based on the size of the predicted phase error, the rate of change, and the dynamic adjustment of acceleration.

[0014] In step S2, the following steps are included:

[0015] S2-1: Input variable error Error change rate With changing acceleration Fuzzy into three variables: x1, x2, x3; the output variable and Fuzzification into multiple linguistic variables;

[0016] S2-2: Get the membership function, then the membership degree can be expressed as:

[0017]

[0018] Among them, a, b, c, d, e, and f are the endpoints of the intervals that define the linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters;

[0019] S2-3: Construct a fuzzy rule table; the constructed fuzzy rule table is specifically:

[0020] Input variable error and error rate of change Combine, and then output the variable and It is jointly determined by the fuzzy variables of both;

[0021] When both inputs are x1 or x3, the output is L; when at least one of the inputs is x2, both will output M; when the inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set.

[0022] S2-4: Perform fuzzy reasoning on the error, error change rate and error acceleration in the control system;

[0023] S2-5: Defuzzify the fuzzy output value;

[0024] Defuzzification converts the fuzzy output obtained by fuzzy reasoning into an accurate value and applies it to actual gain adjustment. The defuzzification method used in the present invention is the centroid method, and the formula is as follows:

[0025]

[0026]

[0027] Where: n is the number of output fuzzy sets; μ i is the membership degree of the output fuzzy set i; is the specific output value corresponding to the fuzzy variable i;

[0028] Assuming the output The fuzzy sets small (S), medium (M), and large (L) correspond to but:

[0029]

[0030] Similarly, assuming the output The fuzzy sets of small (S), medium (M), and large (L) correspond to but:

[0031]

[0032] The above formula can be used to obtain the final accurate output k p (t) and k d (t) and feed it back into the control system.

[0033] The method is used for a LADRC controller used in a power system or a microgrid or distributed generation to improve the robustness of the LADRC controller under dynamic disturbances, system nonlinearity, and complex conditions.

[0034] Compared with the prior art, the present invention has the following technical effects:

[0035] 1) The present invention solves the problem that traditional control methods are difficult to cope with dynamic disturbances, system nonlinearity, and accurate tracking of phase errors in complex environments by combining linear active disturbance rejection control (LADRC) and fuzzy gain optimization methods. Specifically, the phase error feedback characteristics of the LADRC control law are used to achieve accurate suppression of the phase error, and the proportional gain and differential gain are dynamically adjusted through fuzzy control, so that the system can adapt to changes in error, error change rate, and acceleration in real time. This method significantly improves the adaptability, anti-disturbance capability, and robustness of the controller, especially in nonlinear systems or frequent disturbance scenarios, showing excellent stability and tracking accuracy.

[0036] 2) The present invention can significantly improve the phase tracking performance and response speed of the system in a dynamic environment, ensuring the excellent stability of the system under grid frequency fluctuations or external disturbances. In addition, by fuzzifying the input variables and defuzzifying based on the centroid method, this method reduces the dependence on accurate modeling of the system and enhances the adaptability to complex disturbances and uncertain environments. Compared with traditional methods, this method shows higher control effects and application value in terms of accurate phase tracking, dynamic gain optimization, and coping with complex nonlinear disturbances. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The present invention will be further described below in conjunction with the accompanying drawings and embodiments:

[0038] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0039] A method for optimizing the gain of a LADRC controller comprises the following steps:

[0040] Step S1: defining a gain adjustment function of a controller;

[0041] Step S2: setting and optimizing the gain adjustment function;

[0042] In step S1, define the proportional gain k p (t) and the differential gain k d (t).

[0043] In step S1, first define the phase error at the current moment Phase error rate of change And the phase error change acceleration Then define the proportional gain k p (t) and the differential gain k d (t) is:

[0044]

[0045] in, and It is a gain adjustment function that is adjusted in real time based on the size of the predicted phase error, the rate of change, and the dynamic adjustment of acceleration.

[0046] In step S2, the following steps are included:

[0047] S2-1: Input variable error Error change rate With changing acceleration Fuzzy into three variables: x1, x2, x3; the output variable and Fuzzification into multiple linguistic variables;

[0048] S2-2: Get the membership function, then the membership degree can be expressed as:

[0049]

[0050] Among them, a, b, c, d, e, and f are the endpoints of the intervals that define the linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters;

[0051] S2-3: Construct a fuzzy rule table; the constructed fuzzy rule table is specifically:

[0052] Input variable error and error rate of change Combine, and then output the variable and It is jointly determined by the fuzzy variables of both;

[0053] When both inputs are x1 or x3, the output is L; when at least one of the inputs is x2, both will output M; when the inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set.

[0054] S2-4: Perform fuzzy reasoning on the error, error change rate and error acceleration in the control system;

[0055] S2-5: Defuzzify the fuzzy output value;

[0056] Defuzzification converts the fuzzy output obtained by fuzzy reasoning into an accurate value and applies it to actual gain adjustment. The defuzzification method used in the present invention is the centroid method, and the formula is as follows:

[0057]

[0058] Where: n is the number of output fuzzy sets; μi is the membership degree of the output fuzzy set i; is the specific output value corresponding to the fuzzy variable i;

[0059] Assuming the output The fuzzy sets small (S), medium (M), and large (L) correspond to but:

[0060]

[0061] Similarly, assuming the output The fuzzy sets of small (S), medium (M), and large (L) correspond to but:

[0062]

[0063] The above formula can be used to obtain the final accurate output k p (t) and k d (t) and feed it back into the control system.

[0064] This method combines linear active disturbance rejection control (LADRC) and fuzzy gain optimization methods to solve the problem that the control method is difficult to cope with dynamic disturbances, system nonlinearity, and accurate tracking of phase errors in complex environments. Specifically, the phase error feedback characteristics of the LADRC control law are used to achieve accurate suppression of the phase error. At the same time, the proportional gain and differential gain are dynamically adjusted through fuzzy control, so that the system can adapt to changes in error, error change rate, and acceleration in real time. This method significantly improves the controller's adaptability, anti-disturbance capability, and robustness, especially in nonlinear systems or frequent disturbance scenarios, showing excellent stability and tracking accuracy.

[0065] This method can significantly improve the phase tracking performance and response speed of the system in a dynamic environment, ensuring excellent stability of the system under grid frequency fluctuations or external disturbances. In addition, by fuzzifying the input variables and defuzzifying based on the centroid method, this method reduces the reliance on accurate system modeling and enhances the adaptability to complex disturbances and uncertain environments. Compared with traditional methods, this method shows higher control effects and application value in terms of accurate phase tracking, dynamic gain optimization, and coping with complex nonlinear disturbances.

[0066] Example:

[0067] An embodiment provided by the inventor assumes that the method is applied to the frequency control of a power system, and the goal is to optimize the stability of the system frequency through the LADRC controller and reduce the impact of disturbances. Assuming that the frequency fluctuation in the power system will be affected by external disturbances (such as load fluctuations, generator failures, etc.) and internal uncertainties, it is hoped that the proportional gain and differential gain can be optimized through the fuzzy control method to ensure the stability and rapid response of the system.

[0068] Assume that the control goal of the system is to adjust the phase error of the power system Error change rate Acceleration with error variation These variables will affect the output of the LADRC controller and therefore need to be adjusted through fuzzy control.

[0069] Initial phase error: (Power system initial frequency deviation)

[0070] Initial phase error change rate: (Frequency Change Rate)

[0071] Initial phase error change acceleration: (acceleration of frequency change)

[0072] Initial gain: k p =0.5, k d =0.1

[0073] Fuzzify the error, error change rate and acceleration into linguistic variables:

[0074] Phase Error Small (S), Medium (M), Large (L)

[0075] Error change rate Decrease (D), Maintain (Z), Increase (I)

[0076] Error change acceleration Decrease (D), No change (Z), Increase (I)

[0077] k p (proportional gain) and k d (Differential gain): Decrease (L), Remain unchanged (M), Increase (H)

[0078] Set up the following membership function intervals:

[0079] Phase Error S(-0.1Hz,-0.05Hz], M(-0.05Hz,0.05Hz], L(0.05Hz,0.1Hz]

[0080] Error change rate D(-0.2Hz / s,-0.1Hz / s], Z(-0.1Hz / s,0.1Hz / s], I(0.1Hz / s,0.2Hz / s]

[0081] Error change acceleration D(-0.05Hz / s 2 ,-0.02Hz / s 2 ], Z(-0.02Hz / s 2 ,0.02Hz / s 2 ], I(0.02Hz / s 2 ,0.05Hz / s 2 ]

[0082] According to the input fuzzy variables Construct a fuzzy rule table. The following is the fuzzy rule table:

[0083] Table 1 Fuzzy rules table

[0084]

[0085]

[0086] When the phase error is large (L), the error change rate To increase (I), the error changes at an acceleration When the proportional gain k increases (I), p and the differential gain k d should all be increased (H).

[0087] When the phase error For the middle (M), the error change rate is neutral (Z), error change acceleration When it is neutral (Z), the proportional gain k p and the differential gain k d All remain unchanged (M).

[0088] Assume that after fuzzy reasoning, the following fuzzy output is obtained:

[0089] k p is inferred to be M (medium gain), k d Inferred as M (medium gain)

[0090] After fuzzy control, the updated proportional gain and derivative gain are fed back to the LADRC controller:

[0091] New gain values:

[0092] k p =0.6, kd =0.15

[0093] The new gain value will be used to calculate the new control signal, thereby adjusting the output of the power system and reducing the frequency error.

[0094] Feedback: Calculate the control signal using the updated gain (assuming the controller output is calculated as:

[0095]

[0096] Control output: u(0) = 0.045

[0097] The updated error is:

[0098] The new error is:

[0099] Second iteration

[0100] Phase Error:

[0101] Phase error change rate:

[0102] Phase error change acceleration:

[0103] Fuzzy input:

[0104] Then k p =M, k d =M

[0105] The new gain value after defuzzification is:

[0106] k p =0.6, k d =0.15

[0107] Control output: u(1) = -0.00375

[0108] The updated error is:

[0109] The new error is:

[0110] The third iteration

[0111] Phase Error:

[0112] Phase error change rate:

[0113] Phase error change acceleration:

[0114] Fuzzy input:

[0115] Then k p =M, k d =M

[0116] The new gain value after defuzzification is:

[0117] k p =0.6, k d =0.15

[0118] Control output: u(2) = 0.0058125

[0119] The updated error is:

[0120] The new error is:

[0121] In summary, after three rounds of iterations, the frequency error of the system was reduced from 0.05Hz to 0.0029375Hz, and the error was significantly reduced. After each feedback, the proportional gain and differential gain were optimized through fuzzy control, so that the system can adjust the control parameters in real time according to the changes in error, error change rate and acceleration, thereby achieving more stable and faster frequency control.

[0122] The gain and error update process for each iteration is as follows:

[0123] Round 1: Error Updated to

[0124] Round 2: Error Updated to

[0125] Round 3: Error Updated to

[0126] It can be seen that although the error rebounded in the second round, the system gradually stabilized through gain optimization, and finally the error was controlled within a smaller range. This shows that this method has good adaptability and robustness to dynamic disturbances and uncertain environments.

[0127] The present invention combines fuzzy control with LADRC to dynamically optimize the gain of the LADRC controller and dynamically adjust the proportional gain and differential gain of the LADRC controller, thereby optimizing the control effect. The method is suitable for application scenarios such as power systems, microgrids and distributed generation.

Claims

1. A method for optimizing the gain of a LADRC controller, characterized in that: The following steps are involved: Step S1: defining a gain adjustment function of a controller; Step S2: setting and optimizing the gain adjustment function; In step S1, define the proportional gain k p (t) and the differential gain k d (t).

2. The method according to claim 1, characterized in that In step S1, first define the phase error at the current moment Phase error rate of change And the phase error change acceleration Then define the proportional gain k p (t) and the differential gain k d (t) is: in, and It is a gain adjustment function that is adjusted in real time based on the size of the predicted phase error, the rate of change, and the dynamic adjustment of acceleration.

3. The method according to claim 1 or 2, characterized in that: In step S2, the following steps are included: S2-1: Input variable error Error change rate With changing acceleration Fuzzy into three variables: x1, x2, x3; the output variable and Fuzzification into multiple linguistic variables; S2-2: Get the membership function, then the membership degree can be expressed as: Among them, a, b, c, d, e, and f are the endpoints of the intervals that define the linguistic variables, and the shape and position of the membership function are determined by these endpoint parameters; S2-3: Construct a fuzzy rule table; the constructed fuzzy rule table is specifically: Input variable error Error change rate Combine, and then output the variable and It is jointly determined by the fuzzy variables of both; When both inputs are x1 or x3, the output is L; when at least one of the inputs is x2, both will output M; when the inputs are different and neither is x2, the output is S; where L refers to the larger output fuzzy set; M refers to the medium output fuzzy set; and S refers to the smaller output fuzzy set. S2-4: Perform fuzzy reasoning on the error, error change rate and error acceleration in the control system; S2-5: defuzzify the fuzzy output value; Defuzzification converts the fuzzy output obtained by fuzzy reasoning into an accurate value and applies it to actual gain adjustment. The defuzzification method used is the centroid method, and the formula is as follows: Where: n is the number of output fuzzy sets; μ i is the membership degree of the output fuzzy set i; is the specific output value corresponding to the fuzzy variable i; Assuming the output Fuzzy sets small-S, medium-M, and large-L correspond to but: Similarly, assuming the output The fuzzy sets small-S, medium-M, and large-L correspond to but: The above formula can be used to obtain the final accurate output k p (t) and k d (t) and feed it back into the control system.

4. The method according to any one of claims 1 to 3, characterized in that: The method is used for a LADRC controller used in a power system or a microgrid or distributed generation to improve the robustness of the LADRC controller under dynamic disturbances, system nonlinearity, and complex conditions.

Images