A fault detection and regulation method for an electric control system of a wind turbine generator

Abstract

The present application relates to the technical field of electric control, in particular to a kind of fault detection regulation and control method of wind turbine electric control system.The method comprises: according to the similarity between the operating parameters of subsystem B in the duration period when subsystem A occurs fault type a and the operating parameters of subsystem B in the duration period when subsystem B occurs fault type b, obtain the probability representation value of subsystem B when subsystem A occurs fault type a and subsystem B occurs fault type b, obtain the time correlation representation value between the fault type a of subsystem A and the fault type b of subsystem B, according to the probability representation value and the time correlation representation value, obtain the fault correlation result between the fault type a of subsystem A and the fault type b of subsystem B, according to fault rule base and fault correlation result, carry out fault detection and regulation and control to wind turbine electric control system.And the present application can improve the timeliness and the effect of regulation and control of wind turbine electric control system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of electric control, in particular to a fault detection and regulation method for an electric control system of a wind turbine generator. BACKGROUND

[0002] At present, in order to ensure the safe and stable operation of the wind turbine generator, improve the power generation efficiency and economic benefits, and promote the intelligent and digital transformation, it is generally necessary to detect and regulate the fault of the electric control system of the wind turbine generator, that is, to detect and regulate the fault of the electric control system of the wind turbine generator, which is crucial to ensure the safe and stable operation of the unit, improve the power generation efficiency and economic benefits, and promote the intelligent and digital transformation.

[0003] In the prior art, the fault detection of a single subsystem in the electric control system of the wind turbine generator is usually realized based on the fault rule library of the electric control system of the wind turbine generator summarized by experts in the relevant field, and the electric control system of the wind turbine generator is regulated based on the fault detection result of the single subsystem. However, the current regulation only considers the fault detection result of the single subsystem, and does not consider the trend of the future failure of other subsystems affected by the failure of the subsystem, thereby causing poor regulation effect and regulation timeliness. Therefore, how to improve the regulation effect and regulation timeliness has become a problem to be solved. SUMMARY

[0004] In order to solve the above problems, the present application provides a fault detection and regulation method for an electric control system of a wind turbine generator, which adopts the following technical solutions:

[0005] One embodiment of the present application provides a fault detection and regulation method for an electric control system of a wind turbine generator, comprising the following steps:

[0006] Obtain the set of fault types that have occurred for each subsystem of the wind turbine generator electric control system to be analyzed;

[0007] For the fault type a in the set of fault types that have occurred for the subsystem A of the wind turbine generator electric control system to be analyzed and the fault type b in the set of fault types that have occurred for the subsystem B, obtain a probability representation value of the subsystem B occurring fault type b when the subsystem A occurs fault type a according to the similarity between the operating parameters of the subsystem B in the duration period of the subsystem A occurring fault type a and the operating parameters of the subsystem B in the duration period of the subsystem B occurring fault type b, obtain a time correlation representation value between the subsystem A occurring fault type a and the subsystem B occurring fault type b according to the start time of the subsystem A occurring fault type a each time and the start time of the subsystem B occurring fault type b each time, and obtain a fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B according to the probability representation value and the time correlation representation value.

[0008] According to the fault rule base and the fault correlation result, the wind turbine electric control system is detected and regulated.

[0009] Beneficial effects: the present application firstly obtains the fault type set corresponding to each subsystem of the wind turbine electric control system to be analyzed; for the fault type a in the fault type set corresponding to the subsystem A of the wind turbine electric control system to be analyzed and the fault type b in the fault type set corresponding to the subsystem B, according to the similarity between the operation parameters of the subsystem B in the duration period of the fault type a of the subsystem A and the operation parameters of the subsystem B in the duration period of the fault type b, the probability representation value of the fault type b of the subsystem B when the fault type a of the subsystem A occurs is obtained, according to the starting time of the fault type a of the subsystem A and the starting time of the fault type b of the subsystem B, the time correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B is obtained, according to the probability representation value and the time correlation representation value, the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B is obtained; according to the fault rule base and the fault correlation result, the wind turbine electric control system is detected and regulated. And when the wind turbine electric control system is regulated, based on the fault correlation result between the fault type of the subsystem and the fault type of other subsystems except the corresponding subsystem, the timeliness and the regulation effect of the wind turbine electric control system regulation can be improved. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, and the advantages thereof, below will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description only show some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0011] Figure 1 The flow chart of the fault detection and regulation method of the wind turbine electric control system of the present application. DETAILED DESCRIPTION

[0012] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art belong to the protection scope of the embodiments of the present application.

[0013] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0014] This embodiment provides a fault detection and control method for a wind turbine electrical control system, which is described in detail as follows:

[0015] like Figure 1 As shown, the fault detection and control method of the wind turbine electric control system includes the following steps:

[0016] Step S001: obtaining a set of fault types corresponding to each subsystem of the wind turbine electrical control system to be analyzed.

[0017] When regulating the wind turbine electrical control system, this embodiment not only refers to the actual faults of the subsystem, but also considers the trend of future faults of the subsystem that has fault correlation with the faulty subsystem, thereby improving the timeliness of regulation and the effect of regulation. The main manifestation of timeliness of regulation is that regulation is carried out in time when there is a trend of fault occurrence but no fault is generated. The main manifestation of the effect of regulation is that the subsystem that has fault correlation with the faulty subsystem is considered during regulation, and considering the subsystem that has fault correlation with the faulty subsystem during regulation can suppress the chain reaction of faults, thereby improving the effect of regulation. In addition, the improvement of regulation effect and timeliness has a positive effect on reducing failure rate, extending equipment service life, reducing maintenance costs, improving system operation stability, system energy efficiency and energy saving level.

[0018] First, in order to facilitate analysis and understanding, the subsequent will be described with the fault detection and regulation of any type and model of wind turbine electric control system as an example, and the wind turbine electric control system belonging to this type and model will be recorded as the target wind turbine electric control system, the target wind turbine electric control system has been running, that is, the wind turbine electric control system described in the subsequent embodiment is the same type and model of wind turbine electric control system, the target wind turbine electric control system in the embodiment includes consistent subsystems; Then the set of all subsystem types contained in the target wind turbine electric control system is recorded as the comprehensive subsystem set, that is, the subsystems in the comprehensive subsystem set are not repeated, and the subsystems in the comprehensive subsystem set include but are not limited to the main control system, the variable pitch system, the converter system, etc., and then all fault types contained in each subsystem in the comprehensive subsystem set are obtained, and the set of fault types that may occur in each subsystem in the comprehensive subsystem set is recorded as the comprehensive fault type set corresponding to the corresponding subsystem; For example, if the fault types that may occur in the variable pitch system in the comprehensive subsystem set include drive motor failure, sensor failure, gearbox failure, and blade bearing failure, then the set of drive motor failure, sensor failure, gearbox failure, and blade bearing failure is the comprehensive fault type set corresponding to the variable pitch system.

[0019] Next, according to the operation history record and fault record of the target wind turbine electric control system, all fault types that have occurred in each subsystem of each target wind turbine electric control system at the current time are obtained and recorded as the occurred fault types of the corresponding subsystem; For any subsystem of any target wind turbine electric control system, the set of all occurred fault types of the subsystem is recorded as the occurred fault type set corresponding to the subsystem of the target wind turbine electric control system; For example, if the variable pitch system of a certain target wind turbine electric control system has occurred 4 times of failure before the current time, the first failure of the variable pitch system is blade bearing failure, the second failure is gearbox failure, the third failure is drive motor failure, and the fourth failure is blade bearing failure, then it can be known that the known fault type set corresponding to the variable pitch system of the target wind turbine electric control system consists of blade bearing failure, drive motor failure and gearbox failure, and the occurred fault type set corresponding to the variable pitch system of the target wind turbine electric control system belongs to the comprehensive fault type set corresponding to the variable pitch system.

[0020] In addition, the master system is responsible for coordinating the work of each subsystem, according to the wind speed, wind direction and other environmental parameters and the operation state of the unit itself, real-time adjustment of the operation strategy of the wind turbine to achieve maximum power capture and safe and stable operation; the pitch system adjusts the pitch angle of the blade to control the absorption of wind energy by the wind turbine, and plays a role in speed limiting protection when the wind speed is too high; the converter system converts the variable amplitude and frequency AC power generated by the wind turbine into constant frequency and constant voltage AC power that meets the requirements of the power grid, and realizes stable connection with the power grid.

[0021] Therefore, the embodiment obtains the failed type set corresponding to each subsystem of the target wind turbine electric control system and the comprehensive subsystem set through the above process, that is, the comprehensive subsystem sets corresponding to different target wind turbine electric control systems are the same; and at present, the failure detection of a single subsystem in the wind turbine electric control system is usually realized based on the failure rule library of the wind turbine electric control system summarized by experts in the relevant field, and the wind turbine electric control system is regulated and controlled based on the failure detection result of the single subsystem; however, only the failure detection result of the single subsystem is considered when regulating and controlling at present, and the future failure trend of other subsystems affected by the failure of the subsystem is not considered, so the embodiment will analyze the failure correlation between the subsystems in the following to reflect the failure trend of another subsystem when any subsystem fails, that is, the failure correlation between the subsystems can reflect whether another subsystem has a trend of failure of a certain type when a certain subsystem fails.

[0022] In addition, since the analysis process of the fault correlation between any two subsystems in the comprehensive subsystem set is the same, in order to facilitate understanding and analysis, the embodiment will analyze the fault correlation between any fault type a in the comprehensive fault type set corresponding to any subsystem A in the comprehensive subsystem set and fault type b in the comprehensive fault type set corresponding to any subsystem B in the comprehensive subsystem set, subsystem A and subsystem B are not the same, any fault type a in the comprehensive fault type set corresponding to subsystem A can be referred to as fault type a of subsystem A, and fault type b in the comprehensive fault type set corresponding to subsystem B can be referred to as fault type b of subsystem B; Since the target wind turbine electric control system of which the fault type a of subsystem A and the fault type b of subsystem B have occurred needs to be selected to analyze the fault correlation between the fault type a of subsystem A and the fault type b of subsystem B, the embodiment needs to first obtain the target wind turbine electric control system in which the fault type a is included in the occurred fault type set corresponding to any subsystem A and the fault type b is included in the occurred fault type set corresponding to subsystem B according to the occurred fault type set corresponding to subsystem A of all target wind turbine electric control systems and the occurred fault type set corresponding to subsystem B of all target wind turbine electric control systems, and record it as the wind turbine electric control system to be analyzed, that is, all data used in the subsequent analysis of the fault correlation between the fault type a of subsystem A and the fault type b of subsystem B are related to the wind turbine electric control system to be analyzed; and for any target wind turbine electric control system, if the fault type a is included in the occurred fault type set corresponding to subsystem A of the target wind turbine electric control system and the fault type b is also included in the occurred fault type set corresponding to subsystem B of the target wind turbine electric control system, the target wind turbine electric control system can be selected as the wind turbine electric control system to be analyzed for obtaining the fault correlation between the fault type a of subsystem A and the fault type b of subsystem B, that is, if the fault correlation between the fault type a of subsystem A and the fault type b of subsystem B is to be obtained, the target wind turbine electric control system in which the fault type a is included in the occurred fault type set corresponding to subsystem A and the fault type b is included in the occurred fault type set corresponding to subsystem B must be selected as the wind turbine electric control system to be analyzed when analyzing the fault correlation between the fault type a of subsystem A and the fault type b of subsystem B.If the embodiment is analyzed as an example of the fault association between the fault type r of any subsystem A in the comprehensive subsystem set and the fault type t of any subsystem B in the comprehensive subsystem set, the target wind turbine electric control system containing the fault type r in the fault type set of subsystem A and the fault type t in the fault type set of subsystem B must be selected as the wind turbine electric control system to be analyzed when obtaining the fault association result between the fault type r of subsystem A and the fault type t of subsystem B.

[0023] Therefore, the wind turbine electric control system to be analyzed and the fault type set corresponding to each subsystem of the wind turbine electric control system to be analyzed when obtaining the fault association result between the fault type a of subsystem A and the fault type b of subsystem B are obtained through the above process.

[0024] In step S002, according to the similarity between the operation parameters of subsystem B in the duration period of the fault type a of subsystem A and the operation parameters of subsystem B in the duration period of the fault type b, the probability representation value of the fault type b of subsystem B when the fault type a of subsystem A occurs is obtained, the time association representation value between the fault type a of subsystem A and the fault type b of subsystem B is obtained according to the starting time of each fault type a of subsystem A and the starting time of each fault type b of subsystem B, and the fault association result between the fault type a of subsystem A and the fault type b of subsystem B is obtained according to the probability representation value and the time association representation value.

[0025] In the embodiment, the specific obtaining process of the fault association result between the fault type a of subsystem A in the comprehensive subsystem set and the fault type b of subsystem B in the comprehensive subsystem set is as follows:

[0026] First, according to the similarity between the running parameters of the subsystem B of the wind turbine electric control system to be analyzed in the duration period when the subsystem A of the wind turbine electric control system to be analyzed has a fault type a and the running parameters of the subsystem B of the wind turbine electric control system to be analyzed in the duration period when the subsystem B has a fault type b, a probability representation value of the subsystem B having the fault type b when the subsystem A has the fault type a is obtained, according to the starting time of the subsystem A having the fault type a each time and the starting time of the subsystem B having the fault type b each time, a time correlation representation value between the subsystem A having the fault type a and the subsystem B having the fault type b is obtained, and according to the probability representation value and the time correlation representation value, a fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B is obtained; and since the subsystem A and the subsystem B are included in all target wind turbine electric control systems in the embodiment, and the wind turbine electric control system to be analyzed belongs to the target wind turbine electric control system, the subsystem A and the subsystem B appearing in the process of obtaining the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B in the embodiment are the subsystem A and the subsystem B of the wind turbine electric control system to be analyzed, and the fault types appearing are the fault types in the set of fault types corresponding to the subsystem A or the subsystem B of the wind turbine electric control system to be analyzed which have occurred.

[0027] In the embodiment, the specific process of obtaining the probability representation value of the subsystem B having the fault type b when the subsystem A has the fault type a according to the similarity between the running parameters of the subsystem B of the wind turbine electric control system to be analyzed in the duration period when the subsystem A of the wind turbine electric control system to be analyzed has a fault type a and the running parameters of the subsystem B of the wind turbine electric control system to be analyzed in the duration period when the subsystem B has a fault type b is as follows:

[0028] First, according to the record of the to-be-analyzed wind turbine electric control system running at the moment, the duration period of each time the subsystem B of the to-be-analyzed wind turbine electric control system fails type b is obtained, and all the second historical time periods are recorded as the second time period set; for example, if the subsystem B of the to-be-analyzed wind turbine electric control system has occurred 2 times of failure type b, and the starting time of the first time of failure type b of the subsystem B of the to-be-analyzed wind turbine electric control system is t1, the ending time of the first time of failure type b is t2, that is, the subsystem B of the to-be-analyzed wind turbine electric control system fails type b from t1 and lasts until t2, the starting time of the second time of failure type b of the subsystem B of the to-be-analyzed wind turbine electric control system is t3, the ending time of the second time of failure type b is t4, that is, the subsystem B of the to-be-analyzed wind turbine electric control system fails type b from t3 and lasts until t4, then the time period from t1 to t2 is the duration period of the first time of failure type b of the subsystem B of the to-be-analyzed wind turbine electric control system, the time period from t3 to t4 is the duration period of the second time of failure type b of the subsystem B of the to-be-analyzed wind turbine electric control system, and the time period from t1 to t2 is a second historical time period, the time period from t3 to t4 is also a second historical time period, that is, the number of times of failure type b of the subsystem B of the to-be-analyzed wind turbine electric control system is consistent with the number of second historical time periods.

[0029] Then, the duration period of each occurrence of the fault type a of the subsystem A of the wind turbine electric control system to be analyzed before the current time is obtained, and all first operation parameter sequences corresponding to the duration period of each occurrence of the fault type a of the subsystem A of the wind turbine electric control system to be analyzed and all second operation parameter sequences corresponding to each second historical time period in the second time period set of the subsystem B of the wind turbine electric control system to be analyzed are obtained. The determination manner of the duration period of each occurrence of the fault type a of the subsystem A of the wind turbine electric control system to be analyzed is the same as the determination manner of the duration period of each occurrence of the fault type b of the subsystem B of the wind turbine electric control system to be analyzed. The parameters in the first operation parameter sequences and the second operation parameter sequences all belong to the subsystem B of the wind turbine electric control system to be analyzed. The number of the first operation parameter sequences corresponding to the duration period of each occurrence of the fault type a of the subsystem A of the wind turbine electric control system to be analyzed is consistent with the number of the second operation parameter sequences corresponding to any second historical time period, and the number of the first operation parameter sequences corresponding to the duration period of each occurrence of the fault type a of the subsystem A of the wind turbine electric control system to be analyzed or the number of the second operation parameter sequences corresponding to any second historical time period is consistent with the type of the operation parameters collected when the subsystem B of the wind turbine electric control system to be analyzed is monitored.

[0030] Then, according to the similarity between the first operation parameter sequences and the second operation parameter sequences belonging to the same operation parameter type, a feature similarity set between each occurrence of the fault type a of the subsystem A of the wind turbine electric control system to be analyzed and the fault type b of the subsystem B is obtained, and a probability representation value of the fault type b of the subsystem B when the fault type a of the subsystem A occurs is obtained according to the feature similarity set.

[0031] In the embodiment, the specific acquisition process of all first operation parameter sequences corresponding to the duration period of each occurrence of fault type a of subsystem A of the wind turbine electric control system to be analyzed and all second operation parameter sequences corresponding to each second historical time period in the second time period set of subsystem B of the wind turbine electric control system to be analyzed is as follows: for the duration period of the i th occurrence of fault type a of subsystem A of the wind turbine electric control system to be analyzed, first, the duration period of the i th occurrence of fault type a of subsystem A of the wind turbine electric control system to be analyzed is recorded as an analyzed time period, then all operation parameters of subsystem B of the wind turbine electric control system to be analyzed collected in the analyzed time period are acquired, and all operation parameters of subsystem B of the wind turbine electric control system to be analyzed collected in the analyzed time period are composed of all operation parameters collected in the analyzed time period and belonging to subsystem B of the wind turbine electric control system to be analyzed, these historical parameters can be directly acquired from a database storing historical data of the wind turbine electric control system to be analyzed, then all time sequence sequences of all operation parameters of the same operation parameter type in all operation parameters of subsystem B of the wind turbine electric control system to be analyzed in the analyzed time period are recorded as first operation parameter sequences corresponding to the duration period of the i th occurrence of fault type a of subsystem A of the wind turbine electric control system to be analyzed. For the p th second historical time period in the second time period set, first, all operation parameters of subsystem B of the wind turbine electric control system to be analyzed collected in the p th second historical time period are acquired, and all operation parameters of subsystem B of the wind turbine electric control system to be analyzed collected in the p th second historical time period are composed of all operation parameters collected in the p th second historical time period and belonging to subsystem B of the wind turbine electric control system to be analyzed, these historical parameters can also be directly acquired from a database storing historical data of the wind turbine electric control system to be analyzed, then all time sequence sequences of all operation parameters of the same operation parameter type in all operation parameters of subsystem B of the wind turbine electric control system to be analyzed in the p th second historical time period are recorded as second operation parameter sequences corresponding to the p th second historical time period. The purpose of acquiring the first operation parameter sequences and the second operation parameter sequences is to analyze the similarity between the operation parameters of subsystem B when subsystem A occurs fault type a and the operation parameters of subsystem B when subsystem B occurs fault type b, so as to facilitate subsequent calculation of probability representation values, and all operation parameters in any operation parameter sequence are of the same operation parameter type.For example, if the subsystem B is a variable pitch system, and the types of the collected operating parameters of the variable pitch system include variable pitch motor current, variable pitch motor voltage, variable pitch motor temperature, then the number of the first operating parameter sequence corresponding to the duration period when the subsystem A of the wind turbine control system fails for the ith time with the type a and the second operating parameter sequence corresponding to the pth second historical time period are all three, including variable pitch motor current sequence, variable pitch motor voltage sequence, and variable pitch motor temperature sequence. In this embodiment, the types of the collected operating parameters of the subsystem generally include electrical parameters, temperature parameters, mechanical parameters, etc. The electrical parameters include current, voltage, frequency, power factor, etc. The temperature parameters include generator winding temperature, generator bearing temperature, cabin temperature, etc. The mechanical parameters include rotational speed, vibration, etc. Moreover, the collected parameters vary with the subsystem.

[0032] In the embodiment, the specific process of obtaining the characteristic similarity set between each failure type a of subsystem A and failure type b of subsystem B of the wind turbine electric control system to be analyzed according to the similarity between the first sequence set and the second sequence set belonging to the same operating parameter type is as follows: obtaining the first sequence set corresponding to each failure type a of subsystem A of the wind turbine electric control system to be analyzed, and the first sequence set corresponding to the ith failure type a of subsystem A of the wind turbine electric control system to be analyzed is a set constructed by all first sequence sets corresponding to the duration period of the ith failure type a of subsystem A of the wind turbine electric control system to be analyzed. For the gth first sequence set in the first sequence set corresponding to the ith failure type a of subsystem A of the wind turbine electric control system to be analyzed, obtaining the second sequence set with the same operating parameter type as the gth first sequence set from all second sequence sets, and denoting the set of the obtained second sequence set with the same operating parameter type as the gth first sequence set as the gth first sequence set subset, obtaining the similarity between the gth first sequence set and each second sequence set in the gth first sequence set subset, denoting the set of the similarity between the gth first sequence set and each second sequence set in the gth first sequence set subset as a to-be-processed set, and taking the average of all similarities in the to-be-processed set as the characteristic similarity corresponding to the gth first sequence set in the first sequence set corresponding to the ith failure type a of subsystem A of the wind turbine electric control system to be analyzed at the previous moment, and the kth data in the to-be-processed set is the similarity between the gth first sequence set and the kth second sequence set in the gth first sequence set subset; then, denoting the set of the characteristic similarity corresponding to all first sequence sets in the first sequence set corresponding to the ith failure type a of subsystem A of the wind turbine electric control system to be analyzed as the characteristic similarity set between the ith failure type a of subsystem A and the failure type b of subsystem B of the wind turbine electric control system to be analyzed.In addition, it needs to be explained that if all the operating parameters in a certain second operating parameter sequence are variable pitch motor voltages, and all the operating parameters in the gth first operating parameter sequence are also variable pitch motor voltages, it indicates that the operating parameter type of the second operating parameter sequence is the same as the operating parameter type of the gth first operating parameter sequence; the similarity between the gth first operating parameter sequence and the vth second operating parameter sequence in the subset of the gth first operating parameter sequence in the embodiment is the result obtained by negatively correlating the DTW distance between the gth first operating parameter sequence and the vth second operating parameter sequence, that is, the similarity between the gth first operating parameter sequence and the vth second operating parameter sequence is exp(-D), where exp() is the exponential function with constant e as the base, and D is the DTW distance between the gth first operating parameter sequence and the vth second operating parameter sequence; and as other real-time methods, other existing sequence similarity measurement methods can also be used to calculate the similarity between two sequences. Moreover, the size of the feature similarity in the feature similarity set calculated above can reflect the probability or possibility of the occurrence of fault type b of subsystem B when fault type a of subsystem A occurs.

[0033] In this embodiment, the specific process of obtaining the probability representation value of subsystem B having fault type b when subsystem A has fault type a is as follows: first, based on all feature similarities in the feature similarity set between subsystem A of the wind turbine electrical control system to be analyzed and subsystem B of the wind turbine electrical control system to be analyzed having fault type b each time, the probability representation value of subsystem B of the wind turbine electrical control system to be analyzed having fault type b each time subsystem A of the wind turbine electrical control system to be analyzed has been obtained, and the possibility representation value set is obtained, and the hth possibility representation value in the possibility representation value set is the probability representation value of subsystem B of the wind turbine electrical control system to be analyzed. When subsystem A of the electric control system of the wind turbine generator set has fault type a for the hth time, the possibility characterization value of subsystem B of the electric control system of the wind turbine generator set to be analyzed having fault type b is calculated; then the mean of the possibility characterization value set is calculated, and the mean of the possibility characterization value set is used as the probability characterization value of subsystem B having fault type b when subsystem A has fault type a; and the larger the probability characterization value, the greater the probability that subsystem B has fault type b or is about to have fault type b when subsystem A has fault type a, and when the probability characterization value of subsystem B having fault type b when subsystem A has fault type a, the greater the fault correlation between subsystem A's fault type a and subsystem B's fault type b. In addition, in specific applications, the implementer needs to set a preset similarity threshold based on the minimum value of the feature similarity value range. For example, in this embodiment, the minimum value of the feature similarity is 0, so this embodiment sets the preset similarity threshold to 0. Moreover, since the larger the value of the feature similarity that is not 0 and the larger the proportion of the feature similarities that are not 0 in the feature similarity set, it can be indicated that when subsystem A has fault type a, the probability that subsystem B has fault type b or is about to have fault type b is greater.

[0034] In the embodiment, the specific acquisition process of the possibility representation value of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time is described, that is, the specific acquisition process of the possibility representation value of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time is as follows: first, the feature similarity set between the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time and the subsystem B of the wind turbine electric control system under analysis occurring fault type b is denoted as set W0, then a set constructed by all feature similarities not less than a preset similarity threshold in set W0 is obtained and denoted as set W1, the ratio of the total number of feature similarities in set W1 to the total number of feature similarities in set W0 is calculated and denoted as a first index value, the mean of all feature similarities in set W1 is calculated and denoted as a second index value, and finally the product of the first index value and the second index value is calculated and denoted as the possibility representation value of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time.

[0035] In the embodiment, the specific calculation expression of the possibility representation value of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time is as follows:

[0036]

[0037] wherein, is the possibility representation value of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time, N1 is the total number of feature similarities in set W1, N0 is the total number of feature similarities in set W0, is the n-th feature similarity in set W1; and since the conditions of triggering faults of the wind turbine electric control systems of the same type and model are the same, when the value of the feature similarity not equal to 0 is larger and when the proportion of the feature similarity not equal to 0 in the feature similarity set is larger, that is, when the possibility representation value of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time is larger, it can not only indicate that the possibility of the subsystem B of the wind turbine electric control system under analysis occurring fault type b when the subsystem A of the wind turbine electric control system under analysis occurring fault type a for the i-th time is larger, but also indicate that the possibility of the subsystem B of any target wind turbine electric control system occurring fault type b or about to occur fault type b when the subsystem A of the target wind turbine electric control system occurring fault type a for the i-th time is larger.

[0038] In this embodiment, the specific process of obtaining the temporal correlation characterization value between the occurrence of fault type a in subsystem A and the occurrence of fault type b in subsystem B is as follows:

[0039] First, obtain the starting time sequence when fault type a occurs in subsystem A of the electric control system of the wind turbine to be analyzed, and the starting time sequence when fault type b occurs in subsystem B of the electric control system of the wind turbine to be analyzed, and record them as the first time sequence and the second time sequence respectively. The time in the first time sequence is composed of the starting time of each time fault type a occurs in subsystem A before the current moment, and the time in the second time sequence is composed of the starting time of each time fault type b occurs in subsystem B before the current moment. For example, if the number of times subsystem A of the electric control system of the wind turbine to be analyzed has fault type a is 3, and the starting time of the first occurrence of fault type a in subsystem A of the electric control system of the wind turbine to be analyzed is T1, the starting time of the second occurrence of fault type a is T2, and the starting time of the third occurrence of fault type a is T3, then the starting time sequence when fault type a occurs in subsystem A of the electric control system of the wind turbine to be analyzed is {T1, T2, T3}, and the method for obtaining the starting time sequence when fault type b occurs in subsystem B of the electric control system of the wind turbine to be analyzed is the same.

[0040] After judging whether the length of the first time sequence is not greater than the length of the second time sequence, if yes, the first time sequence is recorded as a target sequence, and the second time sequence is recorded as a non-target sequence, otherwise, the second time sequence is recorded as a target sequence, and the first time sequence is recorded as a non-target sequence, the length of the time sequence is the total number of times in the corresponding time sequence; then the time difference value of each time in the target sequence is calculated, and the time difference value of each time is inversely function mapped to obtain the mapping representation value of the corresponding time, and the time difference value of the jth time in the target sequence is the time interval between the time closest to the jth time in the non-target sequence and the jth time; then the mean value of the mapping representation value of all times in the target sequence is calculated and recorded as the time correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B, and the smaller the time difference value of each time in the target sequence, the closer the starting time of the fault type a of the subsystem A and the starting time of the fault type b of the subsystem B, and the more similar the fault occurrence events, the greater the fault correlation between the fault type a of the subsystem A and the fault type b of the subsystem B. Since the smaller the time difference value, the greater the time correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B, therefore, when the time correlation representation value is greater, it indicates that the fault correlation between the fault type a of the subsystem A and the fault type b of the subsystem B is greater, and when the fault correlation between the fault type a of the subsystem A and the fault type b of the subsystem B is greater, it indicates that the subsystem B has a greater trend to occur the fault type b in the future when the fault type a of the subsystem A occurs. In addition, the mapping representation value of the xth time in the target sequence is wherein, is the time difference value of the xth time in the target sequence, c1 is a preset constant, and c1 is to prevent the denominator from being 0.

[0041] In the embodiment, the specific process of obtaining the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B according to the probability representation value of the subsystem B occurring the fault type b when the subsystem A occurs the fault type a and the time correlation representation value between the subsystem A occurring the fault type a and the subsystem B occurring the fault type b is as follows: firstly, the product of the probability representation value of the subsystem B occurring the fault type b when the subsystem A occurs the fault type a and the time correlation representation value between the subsystem A occurring the fault type a and the subsystem B occurring the fault type b is calculated, and the product of the probability representation value and the time correlation representation value is normalized by using the normalization function Norm(), and the normalized result is recorded as the fault correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B; then, it is judged whether the fault correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B is greater than the preset fault correlation threshold value, if yes, it is considered that the fault type a of the subsystem A and the fault type b of the subsystem B have the fault correlation, otherwise, it is considered that the fault type a of the subsystem A and the fault type b of the subsystem B do not have the fault correlation. And when the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B is that when the subsystem A occurs the fault type a, the subsystem B will occur the fault type b, that is, when the fault type a of the subsystem A and the fault type b of the subsystem B have the fault correlation, it indicates that the subsystem B has the trend of occurring the fault type b in a short time in the future or the possibility of the subsystem B occurring the fault type b in a short time in the future is larger, and in order to improve the regulation effect and timeliness, the embodiment requires that the fault correlation can be considered when the system is regulated subsequently, that is, the subsystems which do not occur the fault but have the fault correlation also need to be regulated. In addition, the implementer needs to set the preset fault correlation threshold value according to the value range of the fault correlation representation value, experimental statistics and actual situation, for example, the preset fault correlation threshold value is set to 0.6 in the embodiment. In addition, it needs to be explained that the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B obtained above can represent the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B of any target wind turbine electric control system.

[0042] In step S003, the wind turbine electric control system is detected and regulated according to the fault rule library and the fault correlation result.

[0043] The embodiment can obtain the fault association result between each fault type in the comprehensive fault type set corresponding to each subsystem in the comprehensive subsystem set and each fault type in the comprehensive fault type set corresponding to other subsystems except the corresponding subsystem, i.e., for the subsystem C, the fault association result between each fault type in the comprehensive fault type set corresponding to the subsystem C and each fault type in the comprehensive fault type set corresponding to other subsystems except the subsystem C can be obtained according to the above process. After the fault association result is obtained, the fault of the target wind turbine electric control system F to be detected is detected and regulated according to the known fault rule base of the target wind turbine electric control system and the obtained fault association result. For example, first, the target wind turbine electric control system F to be detected at the current time is detected by using the known fault rule base of the target wind turbine electric control system. If only the fault type a of the subsystem A in the target wind turbine electric control system F is detected at this time, and the fault association result shows that the fault type a of the subsystem A has the fault association with the fault type b of the subsystem B and the fault type a of the subsystem A has the fault association with the fault type y of the subsystem Y, the subsystem A, the subsystem B and the subsystem Y all need to be regulated at the current time, i.e., the running state of the subsystem A, the subsystem B and the subsystem Y is regulated to the standard running state or the running parameter of the subsystem A, the subsystem B and the subsystem Y is regulated to the standard running parameter. The process of detecting the fault of the wind turbine electric control system by using the fault rule base in the embodiment is a known technology. As other real-time methods, the implementer can also set the regulation rules according to the actual situation, but when the regulation rules are set, the other subsystems having the fault association with the detected fault subsystem must be considered.

[0044] Thus, the embodiment completes the fault detection and regulation of the wind turbine electric control system. When the wind turbine electric control system is regulated, the actual fault of the subsystem and the trend of the fault of the subsystem having the fault association with the fault subsystem in the future are considered at the same time, so that the timeliness and the regulation effect of the regulation are improved.

[0045] To sum up, the embodiment obtains the fault type set of each subsystem of the wind turbine electric control system to be analyzed; for the fault type a in the fault type set corresponding to the subsystem A of the wind turbine electric control system to be analyzed and the fault type b in the fault type set corresponding to the subsystem B, according to the similarity between the operation parameters of the subsystem B in the duration of the fault type a of the subsystem A and the operation parameters of the subsystem B in the duration of the fault type b, the probability representation value of the fault type b of the subsystem B when the fault type a of the subsystem A occurs is obtained, according to the starting time of the fault type a of the subsystem A and the starting time of the fault type b of the subsystem B, the time correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B is obtained, according to the probability representation value and the time correlation representation value, the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B is obtained; according to the fault rule library and the fault correlation result, the fault detection and regulation of the wind turbine electric control system are carried out. And when the wind turbine electric control system is regulated, based on the fault correlation result between the fault type of the subsystem and the fault type of other subsystems except the corresponding subsystem, the timeliness and the effect of the regulation of the wind turbine electric control system can be improved.

[0046] The above-described embodiments are only used to illustrate the technical solutions of the present application, but not limit the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can be modified, or some technical features can be replaced by equivalent ones; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A fault detection and control method for an electric control system of a wind turbine generator set, characterized in that: The method comprises the following steps: Obtain a set of fault types that have occurred corresponding to each subsystem of the wind turbine electrical control system to be analyzed; For fault type a in the set of fault types that have occurred corresponding to subsystem A of the electric control system of the wind turbine to be analyzed and fault type b in the set of fault types that have occurred corresponding to subsystem B, based on the similarity between the operating parameters of subsystem B during the duration period of subsystem A's fault type a occurrence and the operating parameters of subsystem B during the duration period of subsystem A's fault type b occurrence, a probability representation value of subsystem B's fault type b occurrence when subsystem A's fault type a occurs is obtained; based on the starting time of each occurrence of fault type a in subsystem A and the starting time of each occurrence of fault type b in subsystem B, a time correlation representation value between the occurrence of fault type a in subsystem A and the occurrence of fault type b in subsystem B is obtained; based on the probability representation value and the time correlation representation value, a fault correlation result between the fault type a of subsystem A and the fault type b of subsystem B is obtained; Perform fault detection and control on the wind turbine electrical control system based on the fault rule base and fault correlation results; A method for obtaining a probability representation value of a subsystem B experiencing a fault type b when a subsystem A experiences a fault type a includes: Obtain the duration of each time the fault type b occurs in the subsystem B, and record each duration as a second historical time period, and record the set of all second historical time periods as a second time period set; Acquire all first operating parameter sequences corresponding to the duration period when the fault type a occurs each time in the subsystem A and all second operating parameter sequences corresponding to each second historical time period in the second time period set; Based on the similarity between the first operating parameter sequence and the second operating parameter sequence belonging to the same operating parameter type, a feature similarity set is obtained between each occurrence of fault type a in subsystem A and each occurrence of fault type b in subsystem B, and based on the feature similarity set, a probability representation value of the occurrence of fault type b in subsystem B when fault type a occurs in subsystem A is obtained, where subsystem A and subsystem B are different; A method for obtaining a temporal correlation characterization value between a fault type a occurring in subsystem A and a fault type b occurring in subsystem B includes: Obtain a starting time sequence of each time a fault type a occurs in subsystem A and a starting time sequence of each time a fault type b occurs in subsystem B, and record them as a first time sequence and a second time sequence, respectively. The time in the first time sequence consists of the starting time of each time a fault type a occurs in subsystem A, and the time in the second time sequence consists of the starting time of each time a fault type b occurs in subsystem B. Determine whether the length of the first time series is not greater than the length of the second time series. If so, record the first time series as the target sequence and the second time series as the non-target sequence. Otherwise, record the second time series as the target sequence and the first time series as the non-target sequence. The time difference of each time in the target sequence is calculated, and the time difference of each time is mapped by an inverse function to obtain a mapping representation value of the corresponding time. The time difference of the j-th time in the target sequence is the time interval between the time closest to the j-th time in the non-target sequence and the j-th time; the mean of the mapping representation values ​​of all times in the target sequence is calculated and recorded as the time correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B.

2. A fault detection and control method for a wind turbine electric control system according to claim 1, characterized in that: The method for obtaining the first operating parameter sequence includes: The duration period when the fault type a occurs for the i-th time in the subsystem A is recorded as the time period to be analyzed, all operating parameters of the subsystem B collected during the time period to be analyzed are obtained, and the time series constructed by the operating parameters of the same operating parameter type among all the operating parameters of the subsystem B collected during the time period to be analyzed are all recorded as the first operating parameter sequence corresponding to the duration period when the fault type a occurs for the i-th time in the subsystem A.

3. The fault detection and control method for the wind turbine electric control system according to claim 1, characterized in that: The method for obtaining the second operating parameter sequence includes: For any second historical time period, all operating parameters of the subsystem B collected within the second historical time period are obtained, and the time series constructed by operating parameters of the same operating parameter type among all operating parameters of the subsystem B collected within the second historical time period are recorded as the second operating parameter sequence corresponding to the second historical time period.

4. A fault detection and control method for a wind turbine electric control system according to claim 1, characterized in that: The method for obtaining the feature similarity set between each failure type a of subsystem A and each failure type b of subsystem B includes: Obtain a first sequence set corresponding to each occurrence of fault type a in subsystem A, where the first sequence set corresponding to the i-th occurrence of fault type a in subsystem A is composed of all first operating parameter sequences corresponding to the duration period when the i-th occurrence of fault type a in subsystem A occurs; For the gth first operating parameter sequence in the first sequence set corresponding to the i-th failure type a of the subsystem A, among all second operating parameter sequences, record the set consisting of all second operating parameter sequences with the same operating parameter type as the g-th first operating parameter sequence as a subset of the g-th first operating parameter sequence. Obtain the average of the similarities between the g-th first operating parameter sequence and each second operating parameter sequence in the subset, and use this average as the feature similarity corresponding to the g-th first operating parameter sequence in the first sequence set corresponding to the i-th failure type a of the subsystem A. The set consisting of the feature similarities corresponding to all first operating parameter sequences in the first sequence set corresponding to the i-th fault type a of the subsystem A is recorded as the feature similarity set between the i-th fault type a of the subsystem A and the fault type b of the subsystem B.

5. The fault detection and control method of the wind turbine electric control system according to claim 1, characterized in that: The method for obtaining, based on the feature similarity set, a probability representation value of subsystem B occurring a fault type b when subsystem A occurs a fault type a, includes: Based on all feature similarities in a feature similarity set between each occurrence of fault type a in subsystem A and each occurrence of fault type b in subsystem B that are not less than a preset similarity threshold, a possibility characterization value of subsystem B occurring fault type b when each occurrence of fault type a in subsystem A is obtained, and a possibility characterization value set is obtained, and the mean of the possibility characterization value set is used as the probability characterization value of subsystem B occurring fault type b when subsystem A occurs fault type a; the hth possibility characterization value in the possibility characterization value set is the possibility characterization value of subsystem B occurring fault type b when subsystem A occurs fault type a for the hth time.

6. A fault detection and control method for a wind turbine electrical control system according to claim 5, characterized in that: A method for obtaining a probability representation value of a subsystem B experiencing a fault type b each time a fault type a occurs in subsystem A includes: For the i-th time that fault type a occurs in subsystem A, the feature similarity set between the i-th time that fault type a occurs in subsystem A and the fault type b occurs in subsystem B is recorded as set W0, a set constructed by all feature similarities in set W0 that are not less than a preset similarity threshold is obtained and recorded as set W1, the ratio of the total number of feature similarities in set W1 to the total number of feature similarities in set W0 is recorded as a first index value, the average of all feature similarities in set W1 is recorded as a second index value, and the product of the first index value and the second index value is recorded as a possibility representation value of the occurrence of fault type b in subsystem B when fault type a occurs in subsystem A for the i-th time.

7. The fault detection and control method for the wind turbine electric control system according to claim 1, characterized in that: The method for obtaining a fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B according to the probability representation value and the time correlation representation value includes: A result obtained by normalizing the product of the probability representation value and the time correlation representation value is recorded as the fault correlation representation value between the fault type a of the subsystem A and the fault type b of the subsystem B; Determine whether the fault correlation characterization value is greater than a preset fault correlation threshold value. If so, the existence of fault correlation between the fault type a of the subsystem A and the fault type b of the subsystem B is taken as the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B. Otherwise, the non-fault correlation between the fault type a of the subsystem A and the fault type b of the subsystem B is taken as the fault correlation result between the fault type a of the subsystem A and the fault type b of the subsystem B.

8. A fault detection and control method for a wind turbine electric control system according to claim 7, characterized in that: The fault correlation between the fault type a of the subsystem A and the fault type b of the subsystem B means that when the fault type a occurs in the subsystem A, it is determined that the fault type b occurs in the subsystem B.

Images