Active power control method for wind turbine group based on fuzzy adaptive control strategy

Abstract

The present invention discloses a wind turbine group active power control method based on a fuzzy adaptive control strategy. The method takes each wind turbine in a wind farm as a measurement point, collects real-time data of the wind turbine, and establishes a data model with three dimensions: health, controllability, and theoretical power index of the wind turbine; receives the total active power target instruction of the wind power cluster in the wind farm, and calculates the controller given value data; filters and analyzes the data according to the real-time operation data model of the wind turbine, and calculates the controller feedback data; utilizes fuzzy control rules to modify PI parameters online in real time to meet the real-time optimal adjustment of PI parameters by wind turbines in different states; processes the controller input based on the discrete PID control algorithm according to the set PI control parameters, and finally calculates the wind farm active control quantity; and then constructs the comprehensive operating force index of the controllable wind turbine group in combination with the theoretical power index of the wind turbine group, and issues a control command to each controllable wind turbine group, so as to optimize the active power output of the entire field.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind farms, in particular to a wind turbine group active power control method based on a fuzzy self-adaptive control strategy for economically, efficiently and healthily controlling the energy of a wind turbine group according to the real-time operation state of the wind turbine group. BACKGROUND

[0002] In recent years, wind power in China has developed rapidly, and the installed capacity of wind power has doubled, and the proportion of wind power in the national installed capacity structure has gradually increased. However, at the same time, the problems of low wind energy utilization rate, intermittency, volatility and randomness of wind energy, and low economic operation level of wind farms still exist, which brings great loss to the overall economic benefit of the wind farm.

[0003] In many areas of China, due to the overdevelopment of wind power, the scale of wind power is not coordinated with the grid accommodation, and the grid dispatching operation problem is becoming more and more prominent, and the phenomenon of wind curtailment and power limitation is also more serious.

[0004] These problems have put forward higher requirements for the safety, stability and control performance of the wind turbine group of the wind farm. In the actual operation process of the wind farm, the characteristic parameters and health status of the wind turbine will change with the influence of different working conditions and other interference factors. The traditional open-loop control, the climbing closed-loop control and the conventional PI control technology can only basically achieve the goal of wind farm power tracking control, and cannot comprehensively consider the real-time operation state of the wind turbine and the actual situation of the whole farm, economically, efficiently and healthily control the energy of the wind turbine group, so as to realize the best adjustment. Improper control strategy not only reduces the service life of the wind turbine, brings more safety hazards, but also is not conducive to the stable and economic operation of the wind farm.

[0005] Therefore, it is necessary to design a wind turbine group active power control method based on a fuzzy self-adaptive intelligent control strategy.

[0006] For example, a "wind farm closed-loop PI controller parameter setting method and device" disclosed in Chinese patent literature, its publication number CN110838725A, includes analyzing each unit, identifying based on the given input signal and the corresponding output signal, obtaining a more ideal single machine model, which estimates the wind farm level active power dynamic model based on the approximate strategy of iterative calculation, and selects a single wind turbine generator system (WTGS) unit as the equivalent model of the wind farm; Based on the equivalent model, the unmodeled disturbance of the wind farm is separated, and the fast Fourier transform is used to find the cutoff frequency of the disturbance; Using the given damping ratio and the cutoff frequency of the disturbance, determine the PI controller parameter debugging of the adaptive PI controller containing the time-varying integral element, and improve the power tracking of the wind farm based on the debugged PI controller. However, the above scheme uses PI control technology to achieve the basic goal of wind farm power tracking control, and there is a problem that the implementation operation state of the unit and the actual situation of the whole field cannot be considered comprehensively. SUMMARY

[0007] The present application is to solve the problem of low utilization rate of wind energy resources, great safety hidden danger and difficult stable, efficient and healthy operation of wind turbines in the prior art wind farm, and provides a wind turbine group active power control method based on fuzzy adaptive control strategy, which can real-time regulate and control the active power of wind turbines according to the real-time operation state of the unit and the actual situation of the whole field.

[0008] To achieve the above purpose, the present application adopts the following technical scheme:

[0009] A wind turbine group active power control method based on fuzzy adaptive control strategy, comprising the following steps:

[0010] (1-1) Data measurement, collection and analysis and evaluation of wind turbine state:

[0011] Taking each wind turbine of the wind farm as a measurement point, collecting real-time data of the wind turbine, storing historical data, analyzing and evaluating to establish a data model of three dimensions of wind turbine health degree, controllability and theoretical power index;

[0012] (1-2) Controller input data processing:

[0013] Receive the total active target instruction of the wind power cluster of the wind farm from the AGC system or by user operation, calculate the given value data of the controller, filter and analyze the real-time operation data model of the wind turbine, and calculate the feedback data of the controller;

[0014] (1-3) PI parameter setting of controller:

[0015] According to the fuzzy adaptive strategy model, taking error e and error change ec as inputs, the PI parameters are modified in real time on line by using fuzzy control rules; the PI parameters are adjusted in real time to meet the optimal adjustment of the wind turbine under different states,

[0016] (1-4) Controller output data processing:

[0017] According to the PI control parameters calibrated in (1-3), the controller input is processed based on the discrete PID control algorithm, and the controller output U is finally calculated; the controller output U is the active control quantity of the wind farm, which is used for the active power control process of the wind turbine group;

[0018] (1-5) Active power control of wind turbine:

[0019] According to the U output by the discrete PID controller, combined with the health index H of the wind turbine, the controllability index S, and the theoretical power index F, the controllable wind turbine comprehensive operation force index J is constructed, and the active power command is sequentially distributed from low to high according to the order of the controllable wind turbine comprehensive operation force index J from low to high. According to the order of the controllable wind turbine comprehensive operation force index J from low to high, the active power command is sequentially distributed according to the active power value from low to high in the active power command.

[0020] The discrete PID control algorithm combined with the operation force intelligent distribution algorithm strategy model issues control commands to the controllable wind turbine, so that the active power output of the whole field reaches the optimal. By reasonably arranging the output performance of the wind turbine under different states, the wind resource utilization rate is improved, unnecessary unit loss is reduced, the whole field wind power output can quickly and safely respond to the change of the grid dispatching instruction, the impact on the grid and the loss of power generation are reduced. The present application has the characteristics of prolonging the service life of the unit, saving cost, improving economy, stability, safety and operation efficiency.

[0021] As preferred, the step (1-1) further comprises the following substeps:

[0022] (1-1-1) Evaluate the health index H of the wind turbine:

[0023] The wind turbine health evaluation factors include gear box oil temperature, gear box vibration, gear box bearing temperature, generator bearing temperature, generator vibration, converter temperature, wind turbine yaw angle and other measurement data; the system combines these real-time data with historical data, and evaluates the current health index H of the wind turbine through expert control algorithm simulation fitting, the health index H is represented by 0%-100%;

[0024] (1-1-2) Evaluate the controllability index S of the wind turbine:

[0025] The controllability evaluation method of the wind turbine acquires state information of the wind turbine, analyzes whether the wind turbine is in a controllable state, records an uncontrollable reason if the wind turbine is uncontrollable, and arranges a priority sequence according to the out-of-control reason for a group of uncontrollable wind turbines; if the wind turbine is controllable, records a current operating state of the wind turbine, and arranges a priority sequence according to the state for a group of controllable wind turbines;

[0026] The controllability index S of the wind turbine ranges from 0% to 100%, and S greater than 50% is set as a controllable wind turbine, and the greater the value, the higher the controllability of the wind turbine; S less than 50% is a wind turbine out of control, and the smaller the value, the lower the possibility of the wind turbine returning to a controllable state;

[0027] The sorting principle is based on expert experience rules, and adjustable input parameters are configured for the control algorithm, and the parameters are continuously optimized through self-learning means, and finally the controllability evaluation index S of each wind turbine is obtained;

[0028] (1-1-3) Evaluation of the theoretical power index F of the wind turbine:

[0029] Through the means of acquisition and statistical analysis, the current real-time wind speed V0, the 1-minute average wind speed V1, the 10-minute average wind speed V2 and the h-time period wind power prediction wind speed V3 of the wind turbine are obtained;

[0030] The theoretical power calculation formula according to the resource method is used to calculate the corresponding theoretical power generation capacity RTheoW0, RTheoW1, RTheoW2 and RTheoW3 of the wind turbine;

[0031] The corresponding historical actual power generation capacity HisTheoW0, HisTheoW1, HisTheoW2 and HisTheoW3 are extracted and analyzed from the historical data by using big data and other means;

[0032] The theoretical power index F of each wind turbine is obtained by the foregoing fuzzy comprehensive evaluation method.

[0033] As preferred, step (1-1-1) further comprises the following steps:

[0034] (1-1-1-1) Health state l of single measurement data ij :

[0035] The health state of single measurement data is evaluated by using a relative degradation degree, and the value range is [0, 1];

[0036] For a smaller-is-better model such as a generator temperature, the calculation formula is:

[0037]

[0038] For the larger the better model such as speed, the calculation formula is:

[0039]

[0040] For the intermediate type such as yaw angle, the calculation formula is:

[0041]

[0042] Wherein, l ij is the degradation of the i-th wind turbine j-th parameter, C i is the measured value of the i-th wind turbine, C max is the maximum value of the wind turbine, C min is the minimum value of the wind turbine, C0 is the optimal value of the wind turbine parameter;(1-1-1-2) the comprehensive health status H i of wind turbine measurement data:

[0043] Based on the degradation analysis of all measured parameters, the fuzzy comprehensive evaluation method is used to comprehensively evaluate and analyze the degradation of each measured parameter and the correlation between the parameter and the health status of the wind turbine, and finally the health status H i of the wind turbine is obtained:

[0044]

[0045] Wherein, H i is the comprehensive health status of the i-th wind turbine, l ij is the degradation of the i-th wind turbine j-th parameter, K ij is the importance of the i-th wind turbine j-th parameter to H i evaluation, or called weight; for the i-th unit, the weight set K i = (K i1 , K i2 , K i3 , …, K ij ), parameter state set L i = (l i1 , l i2 , l i3 , …, l ij ); here, the expert experience algorithm model and the historical fault records accumulated by the fault diagnosis system are introduced, and the experience parameters are provided by the experts in the relevant field, combined with the historical records, the reasonable weight is fitted out, and the weight will be continuously optimized in the subsequent operation of the wind turbine, K ij value is between 0 and 1;

[0046] Based on the fuzzy comprehensive evaluation method, the fuzzy relation matrix is applied to the coincidence operation:

[0047] Hi =L i .K i

[0048] Thus, the health status sequence of each wind turbine is obtained.

[0049] Preferably, the step (1-2) of receiving the total active power target instruction of the wind farm wind power cluster from the AGC system or by user operation, and calculating the controller set value data based on the instruction comprises the following steps:

[0050] The accumulated active power of the uncontrollable wind turbine is P′, and the active target instruction of the wind farm is P0. The controller set value data s is obtained. p =P0-P′.

[0051] Preferably, the filtering and analyzing of the wind turbine real-time operation data model in step (1-2) and calculating the controller feedback data based on the data include the following steps:

[0052] The accumulated active power of the fan is P, and the controller feedback data pv=PP′ is obtained.

[0053] Preferably, step (1-3) further comprises the following steps: (1-3-1) k p Setting principle: When the response is in the rising process (e is P), Δk p Take positive, that is, increase k p ; When overshoot occurs (e is N), Δk p Take the negative value, that is, reduce k p ,When the error is near zero (e is Z), there are three cases: When ec is N, the overshoot becomes larger and larger, at this time Δk p Take negative; when ec is Z, in order to reduce the error, Δk p Take positive; when ec is P, the larger the positive error is, the larger Δk p Get right;

[0054] (1-3-2)k i :Use the integral separation strategy, that is, when the error is near zero, Δk i Take positive, otherwise Δk i Take zero;

[0055] (1-3-3) Define the range of system error e and error change rate ec as the domain on the fuzzy set:

[0056] e,ec={−1, 0, 1}

[0057] Its fuzzy subset is e,ec={N,O,P}, and the elements in the subset represent negative, zero, and positive respectively; e,ec,k p , k iAccording to the membership degree of each fuzzy subset, the fuzzy matrix table of PI parameters is designed by using fuzzy synthesis reasoning, and the correction parameter is substituted into the following formula to calculate the PI parameters:

[0058] k p = k p0 + Δk p , k i = k i0 + Δk i

[0059] In the real-time operation process of the wind turbine group, the controller completes the real-time optimal adjustment of the PI parameters by processing the results of the fuzzy logic rules, looking up the table and performing the operation.

[0060] As preferred, step (1-4) further comprises the following steps:

[0061] (1-4-1) According to the PI parameters set by the controller fuzzy adaptive strategy model, the controller output U(k) is obtained based on the discrete PID control algorithm,

[0062] Formula:

[0063]

[0064] In the formula, k is the sampling sequence number; T is the sampling time.

[0065] As preferred, step (1-5) further comprises the following steps:

[0066] (1-5-1) When the controller output U is greater than the theoretical power index F(t) of all controllable wind turbines , then the active power control instruction of a% of the maximum theoretical power index F(t) of each controllable wind turbine is issued, wherein n is the number of all controllable wind turbines;

[0067] (1-5-2) When the controller output U is less than the theoretical power index F(t) of all controllable wind turbines , that is, the active power intelligent distribution control is performed on all controllable wind turbines on site.

[0068] According to the health degree evaluation H of the wind turbine, the theoretical power index F of the wind turbine and the controllability index S of the wind turbine, the uncontrollable wind turbine is removed, and the multi-objective function of the controllable wind turbine comprehensive operation force index J is constructed as follows:

[0069] J = max (aH + bF + cS)

[0070] Wherein, a is the weight coefficient of the health degree index H of the wind turbine, b is the weight coefficient of the theoretical power index F of the wind turbine, and c is the weight coefficient of the controllability index S of the wind turbine.

[0071] The comprehensive operating force J of the controllable wind turbine is generated, and the sorting method is to allocate the active power instructions from low to high in the order of J from low to high.

[0072] The active power distribution instruction of each controllable wind turbine is as follows:

[0073]

[0074] Where n is the number of controllable fans in the entire site.

[0075] The lower J is, the worse the overall operating capacity of the fan is, and the lower the active power command allocated to it is.

[0076] Therefore, the present invention has the following beneficial effects: (1) The present invention reasonably arranges the output performance of wind turbines under different states, improves the utilization rate of wind resources, reduces the loss of power generation, and enables the wind power output of the entire field to quickly and safely respond to changes in grid dispatching instructions, thereby reducing the impact on the grid and the loss of power generation; (2) According to the discrete PID control algorithm, combined with the operating power intelligent allocation algorithm strategy model, control commands are issued to the controllable wind turbines so that the active power processing of the entire field is optimized, which has the characteristics of being able to extend the service life of the wind turbines, save costs and improve economy, stability, safety and operating efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 This is a flow chart of a method for controlling active power of a wind turbine group according to an embodiment of the present invention. DETAILED DESCRIPTION

[0078] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0079] Example:

[0080] like Figure 1 The method for controlling the power of a wind turbine group based on a fuzzy adaptive intelligent control strategy includes the following steps:

[0081] Take a wind farm consisting of 100 2MW wind turbines as an example:

[0082] (1-1) Data measurement, collection, analysis and evaluation of wind turbine status:

[0083] Using each wind turbine in the wind farm as a measurement point, we collected real-time data from all 100 turbines, stored historical data, and analyzed and evaluated it to establish a data model with three dimensions: turbine health index, controllability index, and theoretical power index.

[0084] (1-1-1) Evaluate the health index H of wind turbines:

[0085] The wind turbine health assessment factors include gear box oil temperature, gear box vibration, gear box bearing temperature, generator bearing temperature, generator vibration, converter temperature, wind turbine yaw angle and other measured data; the system collects these real-time data, combines with historical data, and evaluates the current health index H of the wind turbine through expert control algorithm simulation fitting, and the health index H is represented by 0%-100%;

[0086] (1-1-1-1) Health state of single measurement data ij :

[0087] The health state of single measurement data is evaluated by using relative degradation degree, and the value range is [0, 1];

[0088] For the smaller-the-better model such as generator temperature, the calculation formula is:

[0089]

[0090] For the larger-the-better model such as speed, the calculation formula is:

[0091]

[0092] For the intermediate type such as yaw angle, the calculation formula is:

[0093]

[0094] Wherein, l ij is the degradation degree of the jth parameter of the ith wind turbine, C i is the measured value of the ith wind turbine, C max is the maximum value allowed by the wind turbine for the parameter, c min is the minimum value allowed by the wind turbine for the parameter, and C0 is the optimal value of the parameter of the wind turbine. i :

[0095] Based on the degradation degree analysis of all measurement parameters, the fuzzy comprehensive evaluation method is used to comprehensively evaluate and analyze the degradation degree of each measurement parameter and the correlation between the parameter and the health status of the wind turbine, and finally the health state H of the wind turbine is obtained. i :

[0096]

[0097] Wherein, H i is the comprehensive health state of the ith wind turbine, l ij is the degradation degree of the jth parameter of the ith wind turbine, and K ij is the jth parameter of the ith wind turbine.i Importance of evaluation, or called weight; for the ith unit, the weight set K i = (K i1 , K i2 , K i3 , …, K ij ), parameter state set L i = (l i1 , l i2 , l i3 , …, l ij ); the expert experience algorithm model and the historical fault records accumulated by the fault diagnosis system are introduced here, the experience parameters are provided by the experts in the relevant field, the reasonable weight is fitted out combined with the historical records, the weight will also be continuously optimized in the subsequent operation of the wind turbine, K ij value between 0 and 1;

[0098] Based on the fuzzy comprehensive evaluation method, the coincidence operation of the fuzzy relation matrix is applied:

[0099] H i = L i .K i

[0100] Thus the sequence of the health degree of each wind turbine can be obtained.

[0101] (1-1-2) Evaluate the controllability index S of the wind turbine:

[0102] The controllability evaluation method of the wind turbine collects the state information of the wind turbine, analyzes whether the unit is in a controllable state; if the unit is uncontrollable, record the uncontrollable reason, and the uncontrollable wind turbine group is arranged in priority sequence according to the out-of-control reason; if it can be controlled, record the current operation state of the wind turbine, and the controllable wind turbine group is arranged in priority sequence according to the state;

[0103] The controllability index S of the wind turbine is in the range of 0%-100%, and S greater than 50% is set as a controllable wind turbine, and the larger the value represents the higher the controllability of the wind turbine; S less than 50% is a wind turbine out of control, and the smaller the value represents the lower the possibility of the wind turbine returning to a controllable state;

[0104] The sorting principle is based on the expert experience rule, the adjustable input parameters are configured for the control algorithm, the parameters are continuously optimized through self-learning means, and finally the controllability evaluation index S of each unit is obtained;

[0105] (1-1-3) Evaluate the theoretical power index F of the wind turbine:

[0106] The current real-time wind speed V0, 1-minute average wind speed V1, 10-minute average wind speed V2 and h-time period wind power prediction wind speed V3 of the wind turbine are obtained by means of collection and statistical analysis;

[0107] The theoretical power generation capacity RTheoW0, RTheoW1, RTheoW2 and RTheoW3 corresponding to the wind turbine are respectively calculated according to the theoretical power calculation formula of the resource method;

[0108] The corresponding historical actual power generation capacity HisTheoW0, HisTheoW1, HisTheoW2 and HisTheoW3 are extracted and analyzed from the historical data by means of big data and other means;

[0109] The theoretical power index F of each unit is obtained by the foregoing fuzzy comprehensive evaluation method;

[0110] The three state sequences of the wind turbine, i.e. the health index H, the controllability index S and the theoretical power index F of the wind turbine, are obtained by the foregoing steps.

[0111] (1-2) Controller input data processing:

[0112] The total active target instruction of the wind farm wind power cluster is received from the AGC system or by user operation, and the controller given value data is calculated therefrom, and the controller feedback data is calculated from the real-time operation data model of the wind turbine.

[0113] (1-2-1) Given value sp and feedback value pv calculation:

[0114] The active power of the uncontrollable wind turbine is accumulated as P', the wind farm active target instruction is P0, and sp=P0-P' is obtained. The active power of the wind turbine is accumulated as P, and pv=P-P' is obtained. The foregoing sp and pv are the controller input data.

[0115] (1-3) Controller PI parameter tuning:

[0116] According to the fuzzy adaptive strategy model, the error e (e=sp-pv) and the error change ec (ec(k)=e(k)-e(k-1)) are taken as inputs, and the PI parameters are modified in real time on line by using the fuzzy control rule to meet the real-time optimal adjustment of the PI parameters of the wind turbine under different states.

[0117] As a preferred embodiment, the initial Kp=0.5 and Ki=0.15 are set.

[0118] (1-3-1) k p Tuning principle: when the response is in the rising process, i.e. e is P, Δk pTake positive, that is, increase k p ; when overshoot, that is, e is N, Δk p Take negative, that is, reduce k p When the error is near zero, that is, e is Z, there are three cases: ec is N, the overshoot is getting bigger and bigger, at this time Δk p Take negative; ec is Z, in order to reduce the error, Δk p Take positive; ec is P, the positive error is getting bigger and bigger, Δk p Take positive.

[0119] (1-3-2) k i : integral separation strategy, that is, when the error is near zero, Δk i Take positive, otherwise Δk i Take zero.

[0120] (1-3-3) the system error e and the error change rate ec change range is defined as the domain of fuzzy set:

[0121] e, ec = {-1, 0, 1}

[0122] Its fuzzy subsets are e, ec = {N, O, P}, the elements in the subset represent negative, zero, positive respectively. e, ec, k p , k i Subject to normal distribution, the membership degree of each fuzzy subset is obtained, according to the membership degree assignment table of each fuzzy subset and the parameter fuzzy control model, the fuzzy matrix table of PI parameter is designed by applying fuzzy synthesis reasoning, and the correction parameter is found out to be substituted into the following formula.

[0123] k p = k p0 + Δk p , k i = k i0 + Δk i

[0124] In the real-time operation process of wind turbine group, the controller completes the real-time optimal adjustment of PI parameters through the result processing, table lookup and operation of fuzzy logic rules.

[0125] (1-4) controller output data processing:

[0126] According to the PI control parameters adjusted in (1-3), the controller input is processed based on discrete PID control algorithm, and the active control control quantity of wind farm is finally calculated, which is used for active power control process of wind turbine group.

[0127] (1-4-1) according to the PI parameters adjusted by the fuzzy adaptive strategy model of the controller, the controller output U(k) is obtained based on discrete PID control algorithm,

[0128] Formula:

[0129]

[0130] In the formula, k is the sampling number; T is the sampling time, and the time T of one cycle can be selected as 1 second.

[0131] (1-5) Active power control of wind turbine generator:

[0132] According to the output U of the discrete PID controller, combined with the health index H of the wind turbine generator, the controllability index S and the theoretical power index F, the controllable wind turbine generator comprehensive operation force index J is constructed, and the active power command from low to high is distributed in turn according to the order of the controllable wind turbine generator comprehensive operation force index from low to high. Thus, the total field active power output reaches the optimum.

[0133] (1-5-1) When the controller output U is greater than the total field controllable wind turbine theoretical power index , where n is the number of total field controllable wind turbines, then the active power control command of a% of the maximum theoretical power index F(t) of each controllable wind turbine is issued, and according to the actual operation of the wind farm, a% can be selected as 10%;

[0134] (1-5-2) When the controller output U is less than the total field controllable wind turbine theoretical power index , that is, the active power intelligent distribution control is performed on all controllable wind turbines in the field.

[0135] Considering the health index H of the wind turbine generator, the theoretical power index F of the wind turbine generator and the controllability index S of the wind turbine generator, the uncontrollable wind turbine generator is eliminated, and the multi-objective function of the controllable wind turbine generator comprehensive operation force index J is constructed as follows:

[0136] J = max (aH + bF + cS)

[0137] Wherein a is the weight coefficient of the health index H of the wind turbine generator, b is the weight coefficient of the theoretical power index F of the wind turbine generator, and c is the weight coefficient of the controllability index S of the wind turbine generator.

[0138] Thus, the controllable wind turbine generator comprehensive operation force J is generated, and the sorting mode is according to the order of J from low to high. The lower J is, the worse the comprehensive operation force of the wind turbine is, and the lower the active power command allocated is.

[0139] Therefore, the active power distribution command of each controllable wind turbine generator is as follows:

[0140]

[0141] Wherein n is the number of the whole field controllable fan. By reasonably arranging the output performance of the wind turbine under different states, the wind resource utilization rate is improved, unnecessary unit loss is reduced, the whole field wind power output can quickly and safely respond to the change of the grid dispatching instruction, the impact on the power grid and the loss of power generation are reduced. The application has the characteristics of prolonging the service life of the unit, saving cost, improving economy, stability, safety and operation efficiency.

[0142] The specific embodiments described herein merely exemplify the spirit of the present application. Those skilled in the art of the present application can make various modifications or supplements to the described specific embodiments or use similar ways to replace them, but will not deviate from the spirit of the present application or exceed the scope defined by the appended claims.

[0143] Although the terms such as fuzzy adaptive strategy model, setting principle, sequence, accumulation, fuzzy relationship matrix are used more in this paper, but it does not exclude the possibility of using other terms. The use of these terms is only to facilitate the description and explanation of the essence of the application; any kind of additional limitation is contrary to the spirit of the application.

Claims

1. A wind turbine group active power control method based on fuzzy adaptive control strategy, characterized in that: The following steps are included: (1-1) Data measurement, collection, analysis and evaluation of wind turbine status: Taking each wind turbine in the wind farm as a measurement point, real-time data of the wind turbine is collected, historical data is stored, and analysis and evaluation are conducted to establish a real-time operation data model of the wind turbine in three dimensions: wind turbine health index, controllability index, and theoretical power index; (1-2) Discrete PID controller input data processing: Receive the total active power target instruction of the wind farm wind power cluster from the AGC system or through user operation, and calculate the discrete PID controller set value data; filter and analyze the real-time operation data model of the wind turbine group to calculate the discrete PID controller feedback data; the set value data and feedback data are the input data of the discrete PID controller; (1-3) Discrete PID controller PI parameter tuning: Based on the fuzzy adaptive strategy model, the error e between the total active power target command of the wind farm wind power cluster and the real-time active power of the wind turbine units, as well as the change ec of the error e between adjacent sampling numbers k and k-1, are used as inputs. The fuzzy control rules are used to modify the PI parameters in real time online. (1-4) Discrete PID controller output data processing: According to the PI control parameters adjusted in (1-3), the discrete PID controller input is processed based on the discrete PID control algorithm, and the discrete PID controller output U is finally calculated; (1-5) Active power control of wind turbines: According to the output U of the discrete PID controller, combined with the wind turbine health index H, controllability index S, and theoretical power index F, the comprehensive operating force index J of the controllable wind turbine is constructed. Each controllable wind turbine corresponds to a comprehensive operating force index J. The active power command is allocated to the corresponding controllable wind turbine in the order of the comprehensive operating force index J from low to high and the active power value in the active power command from low to high.

2. The method for controlling active power of a wind turbine group based on fuzzy adaptive control strategy according to claim 1 is characterized in that: The step (1-1) further includes the following subdivision steps: (1-1-1) Evaluate the health index H of wind turbines: Wind turbine health assessment factors include gearbox oil temperature, gearbox vibration, gearbox bearing temperature, generator bearing temperature, generator vibration, converter temperature, and wind turbine yaw angle measurement data. The system collects this real-time data, combines it with historical data, and uses expert control algorithm simulation and fitting to evaluate the wind turbine's current health index H, which is expressed on a scale of 0%-100%. (1-1-2) Evaluate the controllability index S of wind turbines: The wind turbine controllability assessment method collects wind turbine status information and analyzes whether the turbine is in a controllable state. If the turbine is uncontrollable, the cause of the uncontrollability is recorded, and the uncontrollable wind turbine groups are prioritized according to the cause of the uncontrollability. If the turbine is controllable, the current wind turbine operating status is recorded, and the controllable wind turbine groups are prioritized according to the status. The wind turbine controllability index S ranges from 0% to 100%. When S is greater than 50%, the wind turbine can be controlled. The larger the value, the higher the controllability of the wind turbine. When S is less than 50%, the wind turbine is out of control. The smaller the value, the lower the possibility of the wind turbine returning to a controllable state. The sorting principle is based on the expert experience rule, and the control algorithm is configured with adjustable input parameters. Through self-learning, the parameters are continuously optimized to finally obtain the controllability evaluation index S of each unit. (1-1-3) Evaluate the theoretical power index F of wind turbines: By means of data collection and statistical analysis, the current real-time wind speed V0, 1-minute average wind speed V1, 10-minute average wind speed V2 of the wind turbine and the wind power forecast wind speed V3 within the h time period are obtained; The theoretical power generation capacities RTheoW0, RTheoW1, RTheoW2 and RTheoW3 corresponding to the wind turbines are calculated respectively according to the theoretical power calculation formula of the resource method; Using big data technology to extract and analyze historical data to obtain the corresponding historical actual power generation capacity HisTheoW0, HisTheoW1, HisTheoW2, HisTheoW3; The theoretical power index F of each unit is evaluated and calculated using the fuzzy comprehensive evaluation method.

3. The method for controlling active power of a wind turbine group based on a fuzzy adaptive control strategy according to claim 2 is characterized in that: Step (1-1-1) further includes the following steps: (1-1-1-1) Health status of a single measurement data ij : The health status of a single measurement data is evaluated using relative degradation, with a value range of [0,1]; For the smaller the better model, the calculation formula is: For the larger the better model, the calculation formula is: For the intermediate model, the calculation formula is: Among them, l ij is the degradation degree of the jth parameter of the i-th wind turbine, C i is the measured value of the i-th wind turbine, C max is the maximum value allowed by the jth parameter of the wind turbine, C min is the minimum value allowed for the jth parameter of the wind turbine generator set, and C0 is the optimal value for the jth parameter of the wind turbine generator set; (1-1-1-2) Comprehensive health status H of wind turbine measurement data i : Based on the degradation analysis of all measured parameters, the fuzzy comprehensive evaluation method is used to conduct a comprehensive evaluation and analysis of the degradation degree of each measured parameter and the correlation between the parameter and the health status of the wind turbine, and finally the health status H of the wind turbine is obtained. i : Among them, H i is the comprehensive health status of the i-th wind turbine, l ij is the degradation degree of the jth parameter of the i-th wind turbine, K ij H is the jth parameter pair of the i-th wind turbine i The importance of the assessment, or called weight; for the i-th unit, the weight set K i =(K i1 ,K i2 ,K i3 ,…,K ij ), parameter state set L i =(l i1 ,l i2 ,l i3 ,…,l ij ); Here, we introduce the expert experience algorithm model and the historical fault records accumulated by the fault diagnosis system. The relevant field experts provide experience parameters, and combine the historical records to fit the reasonable weight. The weight will also be continuously optimized in the subsequent operation of the wind turbine. ij The value is between 0 and 1; Based on the fuzzy comprehensive evaluation method, the composite operation of the fuzzy relationship matrix is ​​applied: H i =L i .K i Thus, the health status sequence of each wind turbine is obtained.

4. The method for controlling active power of a wind turbine group based on a fuzzy adaptive control strategy according to claim 3 is characterized in that: The step (1-2) of receiving the total active power target instruction of the wind farm wind power cluster from the AGC system or by user operation and calculating the discrete PID controller given value data based on the instruction includes the following steps: The accumulated active power of the uncontrollable wind turbines is P', and the active target instruction of the wind farm is P0. The discrete PID controller given value data sp = P0-P' is obtained.

5. The method for controlling active power of a wind turbine group based on a fuzzy adaptive control strategy according to claim 4 is characterized in that: The filtering and analysis of the real-time operation data model of the wind turbine generator described in step (1-2) and the calculation of discrete PID controller feedback data based on the data include the following steps: The accumulated active power of the wind turbine is P, and the discrete PID controller feedback data pv=PP′ is obtained.

6. The method for controlling active power of a wind turbine group based on a fuzzy adaptive control strategy according to claim 1 is characterized in that: Steps (1-3) also include the following steps: (1-3-1)k p Setting principle: When the response is in the rising process, e is P, Δk p Take positive, that is, increase k p ; When overshoot occurs, e is N, Δk p Take the negative value, that is, reduce k p ,When the error is near zero, e is Z, there are three cases: When ec is N, the overshoot becomes larger and larger, at this time Δk p Take negative; when ec is Z, in order to reduce the error, Δk p Take positive; when ec is P, the larger the positive error is, the larger Δk p Get right; (1-3-2)k i :Use the integral separation strategy, that is, when the error is near zero, Δk i Take positive, otherwise Δk i Take zero; (1-3-3) Define the range of system error e and error change rate ec as the domain on the fuzzy set: e,ec={-1,0,1} Its fuzzy subset is e,ec={N,O,P}, and the elements in the subset represent negative, zero, and positive respectively; e,ec,k p , k i Obeying the normal distribution, the membership of each fuzzy subset is obtained. According to the membership assignment table of each fuzzy subset and the fuzzy control model of each parameter, the fuzzy matrix table of PI parameters is designed by fuzzy synthesis reasoning, and the correction parameters are found and substituted into the following formula for calculation; k p =k p0 +Δk p ,k i =k i0 +Δk i During the real-time operation of the wind turbine group, the controller completes the real-time optimal adjustment of the PI parameters by processing the results of fuzzy logic rules, looking up tables and calculating.

7. The method for controlling active power of a wind turbine group based on a fuzzy adaptive control strategy according to claim 1 is characterized in that: Steps (1-4) also include the following steps: (1-4-1) Based on the PI parameters adjusted by the discrete PID controller fuzzy adaptive strategy model, the controller output U(k) is obtained based on the discrete PID control algorithm. formula: Where k is the sampling number; T is the sampling time, e(k) is the error at sampling number k, and e(j) is the error at sampling number j, j = 0, ..., k.

8. The method for controlling active power of a wind turbine group based on a fuzzy adaptive control strategy according to claim 1 is characterized in that: Steps (1-5) also include the following steps: (1-5-1) When the controller output U is greater than the theoretical power index of the controllable wind turbines in the entire field When , an active power control instruction is issued to each controllable wind turbine generator set, which increases its maximum theoretical power index F(t) by a%, where n is the number of controllable wind turbine generator sets in the entire field; (1-5-2) When the controller output U is less than the theoretical power index of the controllable wind turbines in the entire field At this time, the active power intelligent distribution control is performed on all controllable wind turbines on site; According to the health evaluation H of wind turbines, the theoretical power index F of wind turbines, and the controllability index S of wind turbines, the uncontrollable wind turbines are eliminated and the multi-objective function of the comprehensive operating force index J of controllable wind turbines is constructed as follows: J=max(aH+bF+cS) Among them, a is the weight coefficient of the wind turbine health index H, b is the weight coefficient of the wind turbine theoretical power index F, and c is the weight coefficient of the wind turbine controllability index S; The comprehensive operating force index J of the controllable wind turbine is generated, and the sorting method is to assign active power instructions from low to high in the order of J. The active power distribution instruction of each controllable wind turbine is as follows: Where n is the number of controllable wind turbines in the entire field.

Images