A method and system for controlling an offshore wind farm

Abstract

The application discloses an offshore wind farm control method and system, relates to the technical field of wind power generation control, and comprises the following steps: collecting a three-dimensional environment parameter vector through a sensor array arranged on a wind turbine generator; inputting real-time environment parameters into a pre-trained classification model and matching to a plurality of preset environment operating states; constructing a Markov state transition probability matrix based on historical operation data and recording the environment parameter change rate characteristics at each state transition; determining the adaptation degree of the preset control parameters of the wind turbine generator in each environment operating state to other environment operating states; fusing to generate a hybrid prediction result; and triggering a control strategy according to the comprehensive judgment of the prediction result and the adaptation degree evaluation. The application has the advantages that the dynamic Markov state transition prediction and the cross-state adaptation degree evaluation are combined, and the self-adaptability of the offshore wind farm to complex sea conditions is significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wind power generation control, in particular to a kind of offshore wind farm control method and system. BACKGROUND

[0002] As an important part of clean energy, offshore wind power is often faced with multiple challenges such as sudden wind speed change, wave impact and salt spray corrosion due to its complex and variable operating environment. Traditional control methods rely on single environmental parameter or static threshold for decision-making, which is difficult to dynamically adapt to comprehensive environmental changes, resulting in fluctuation of power generation efficiency, aggravation of equipment wear and tear, and even safety hazards in extreme working conditions. Especially in severe sea conditions, the existing technology lacks the ability to analyze multi-dimensional environmental parameters cooperatively, making it difficult to predict the state transition trend in time, resulting in lagging control strategy and restricting the reliability and economy of offshore wind farm.

[0003] In the current scheme, environmental state recognition is mostly based on fixed rules or single machine learning model, and the time sequence correlation and state transition law of historical operation data are not fully explored. At the same time, the adjustment of control parameters often only considers the current working condition without considering the future environmental evolution trend and cross-state adaptation degree evaluation, which may cause frequent start-stop or suboptimal control. In addition, the traditional Markov prediction model has low parameter updating efficiency in dynamic environment, and cannot effectively integrate real-time data and historical state transition features, resulting in limited prediction accuracy. There is an urgent need for a closed-loop control system that integrates multi-source environmental perception, intelligent state classification and dynamic prediction decision-making to improve the adaptive operation ability and risk prevention level in complex sea conditions SUMMARY

[0004] To solve the above technical problems, a kind of offshore wind farm control method and system are provided, which solves the problem that the existing technology lacks the ability to analyze multi-dimensional environmental parameters cooperatively, making it difficult to predict the state transition trend in time, resulting in lagging control strategy and restricting the reliability and economy of offshore wind farm.

[0005] To achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0006] A kind of offshore wind farm control method, comprising:

[0007] Collect real-time wind speed, significant wave height and salt spray concentration data through sensor array deployed on wind turbine generator set to form a three-dimensional environmental parameter vector;

[0008] Input real-time environmental parameters into a pre-trained classification model and match them to a plurality of preset environmental operating states, including normal power generation mode, storm warning mode and extreme shutdown mode;

[0009] Construct a Markov state transition probability matrix based on historical operation data, and record the environmental parameter change rate characteristics at each state transition;

[0010] Determine the adaptation degree of the wind turbine preset control parameters in each environmental operating state to other environmental operating states based on the wind turbine preset control parameters in each environmental operating state;

[0011] Based on the environmental parameter change rate characteristics at each state transition, predict the environmental parameter change in the future period and generate a hybrid prediction result by combining the state transition probability matrix;

[0012] According to the comprehensive judgment of the prediction result and the adaptation degree evaluation, trigger one of the control strategies of maintaining operation, parameter adjustment or emergency shutdown.

[0013] Preferably, the specific training process of the classification model is:

[0014] Set an environmental parameter clustering center vector for each environmental operating state respectively;

[0015] Based on the historical control log of the offshore wind farm, determine a plurality of environmental parameter vectors at the adjustment of each environmental operating state, denoted as sample environmental parameter vectors;

[0016] Based on the plurality of sample environmental parameter vectors and the environmental parameter clustering center vector, determine the environmental parameter boundary value of each environmental operating state;

[0017] Based on the environmental parameter clustering center vector and the environmental parameter boundary value, determine the classification model of the environmental operating state.

[0018] Preferably, the Markov state transition probability matrix is constructed based on historical operation data, and the environmental parameter change rate characteristics at each state transition are recorded, which specifically includes:

[0019] Determine an analysis duration based on the control attributes of the wind turbine;

[0020] Set a sample collection duration, and set the time period closest to the current time within the sample collection duration as the sample collection period;

[0021] Based on the analysis duration, divide the sample collection period into a plurality of analysis periods;

[0022] Determine the environmental state of each analysis period based on the classification model;

[0023] Based on the environmental state of each analysis period, construct a Markov state transition probability matrix, , A Markov state transition probability of a transition from an i-th environmental operating state to a j-th environmental operating state, K is a total number of preset environmental operating states;

[0024] wherein, , A total number of analysis periods in the i-th environmental state, A number of adjacent analysis periods in which the environmental state is changed from the i-th environmental state to the j-th environmental operating state;

[0025] An environmental parameter average value in each analysis period and a change rate of the environmental parameter over time in each analysis period are respectively counted, and a change rate feature of the environmental parameter in the state transition is obtained.

[0026] Preferably, the determining of the adaptation degree of the preset control parameter of the wind turbine under each environmental operating state to other environmental operating states specifically comprises:

[0027] A performance index of the wind turbine under the preset control parameter of the wind turbine under each environmental operating state is determined, and the performance index at least includes yaw delay, power fluctuation rate and blade stress peak value;

[0028] The calculation formula of the adaptation degree is:

[0029]

[0030] In the formula, An adaptation degree of the preset control parameter of the wind turbine under the k-th environmental operating state to the h-th environmental operating state, A total number of performance indexes, An n-th performance index corresponding to the preset control parameter of the wind turbine under the k-th environmental operating state, An n-th performance index corresponding to the preset control parameter of the wind turbine under the h-th environmental operating state, A weight of the n-th performance index.

[0031] Preferably, the predicting of the environmental parameter change in a future period based on the change rate feature of the environmental parameter in the state transition specifically comprises:

[0032] A real-time parameter value of the real-time collected environmental parameter and a real-time change rate of the environmental parameter are based on the environmental parameter average value in each analysis period and the change rate of the environmental parameter over time, a vector distance is solved, and a similarity distance index of the real-time environment and the environment in each analysis period is obtained;

[0033] An environmental operating state in a next analysis period of each analysis period is counted;

[0034] a transition fitting probability of each environment running state in the future analysis duration is calculated based on a similar distance index of the real-time environment and the environment in all analysis periods and an environment running state in a next analysis period of each analysis period;

[0035] Specifically, the calculation formula is:

[0036]

[0037] wherein, is a transition fitting probability of the i-th environment running state in the future analysis duration, is a set of analysis periods in which the i-th environment running state is in the next analysis period, c is the c-th element in the set, E is a set of all analysis periods, o is the o-th element in E, is a maximum value in the similar distance index of the real-time environment and the environment in all analysis periods, is a minimum value in the similar distance index of the real-time environment and the environment in all analysis periods, is a similar distance index of the real-time environment and the c-th environment.

[0038] Preferably, the specific steps of generating a hybrid prediction result by fusing the state transition probability matrix include:

[0039] determining an environment running state at a current time, and determining a Markov state transition probability of transition to each environment running state from the Markov state transition probability matrix based on the environment running state at the current time;

[0040] setting a hybrid weight of the transition fitting probability and the Markov state transition probability respectively, and performing weighted summation on the transition fitting probability and the Markov state transition probability to obtain a hybrid prediction result.

[0041] Preferably, the comprehensive judgment according to the prediction result and the adaptation degree evaluation triggers one of the control strategies of maintaining operation, parameter adjustment or emergency shutdown, specifically including:

[0042] if the prediction probability of the extreme shutdown mode in the hybrid prediction result exceeds 0.5, emergency shutdown is triggered;

[0043] otherwise, the prediction probability of each environment running state in the hybrid prediction result is used as a weight to perform weighted summation on the adaptation degree of the wind turbine preset control parameter and the environment running state under each environment running state to obtain a prediction adaptation degree of the wind turbine preset control parameter and the environment under each environment running state;

[0044] if the environment running state corresponding to the maximum prediction adaptation degree is the same as the current environment running state, maintaining operation is triggered.

[0045] If the environment running state corresponding to the maximum prediction fitness is different from the current environment running state, a parameter adjustment is triggered, and the control parameter of the offshore wind farm is adjusted to the preset control parameter of the wind turbine generator set in the environment running state corresponding to the maximum prediction fitness.

[0046] Further, an offshore wind farm control system is proposed, which is used to implement the offshore wind farm control method as described above, and comprises:

[0047] A sensor array is arranged in the sensor array of the wind turbine generator set, and is used to collect wind speed, effective wave height and salt mist concentration data in real time.

[0048] A data processing module is used to integrate the collected real-time data into a three-dimensional environmental parameter vector.

[0049] A state classification module is internally provided with a pre-trained classification model, and the three-dimensional environmental parameter vector is matched to a plurality of preset environment running states, including a normal power generation mode, a storm warning mode and an extreme shutdown mode.

[0050] A Markov analysis module is used to construct a state transition probability matrix based on historical operation data, and store the environmental parameter change rate characteristics at each state transition.

[0051] An adaptation degree evaluation module is used to calculate the control adaptation degree of each environment running state to other states according to the preset control parameter of each environment running state.

[0052] An environment prediction module is used to generate a hybrid prediction result by combining the real-time environmental parameter change rate and the state transition probability matrix.

[0053] A control strategy execution module is used to trigger one of a maintenance operation, a parameter adjustment or an emergency shutdown according to the hybrid prediction result and the adaptation degree evaluation result.

[0054] Compared with the prior art, the offshore wind farm control method and system have the following advantages:

[0055] The application significantly improves the self-adaptive ability of the offshore wind farm to complex sea conditions by fusing multi-source environmental parameters for real-time perception and intelligent state classification, combining dynamic Markov state transition prediction and cross-state adaptation evaluation. The multi-dimensional environmental parameter collaborative analysis and hybrid prediction mechanism effectively improves the extreme working condition warning accuracy, avoiding the misjudgment or response lag caused by traditional static threshold; by predicting the environmental evolution trend and the adaptability of cross-state control parameters, the generator set operation strategy is dynamically optimized, reducing unnecessary downtime and frequent parameter adjustment, while ensuring equipment safety and improving power generation efficiency. In addition, the fusion prediction model based on historical data and real-time change rate enhances the robustness of the system to sudden environmental disturbances, achieving power generation loss reduction and equipment maintenance cost reduction. BRIEF DESCRIPTION OF DRAWINGS

[0056] Figure 1 A flow chart of the offshore wind farm control method proposed in the present scheme is provided.

[0057] Figure 2 A flow chart of the specific training method of the classification model proposed in the present scheme is provided.

[0058] Figure 3 A flow chart of the method for constructing the Markov state transition probability matrix and recording the environmental parameter change rate characteristics of each state transition proposed in the present scheme is provided.

[0059] Figure 4 A flow chart of the method for predicting the environmental parameter change of the future period proposed in the present scheme is provided.

[0060] Figure 5 A flow chart of the method for generating a hybrid prediction result proposed in the present scheme is provided.

[0061] Figure 6 An architecture diagram of the electronic device in the present scheme is provided.

[0062] Figure 7 A structure diagram of the computer readable storage medium in the present scheme is provided.

[0063] The figure numbers are:

[0064] 500-electronic device; 501-bus; 502-CPU; 503-ROM; 504-RAM; 505-communication port; 506-input / output component; 507-hard disk; 508-user interface; 600-computer readable storage medium. DETAILED DESCRIPTION

[0065] The following description is used to disclose the present application so that those skilled in the art can implement the present application. The preferred embodiments in the following description are only as examples, and other obvious modifications can be made by those skilled in the art.

[0066] Referring to Figure 1 An offshore wind farm control method is shown, comprising:

[0067] Real-time wind speed, significant wave height, and salt spray concentration data are collected by a sensor array deployed on a wind turbine generator unit to form a three-dimensional environmental parameter vector;

[0068] Through real-time collaborative monitoring of multi-dimensional environmental parameters, the limitations of traditional single parameter perception are broken through, and core environmental factors such as wind speed, wave, and salt spray are comprehensively integrated to accurately represent the dynamic characteristics of complex sea conditions, providing multi-dimensional data support for subsequent state classification and risk assessment, and significantly reducing the risk of abnormal equipment loss caused by environmental factor coupling;

[0069] The real-time environmental parameters are input into a pre-trained classification model and matched to a plurality of preset environmental operating states, including at least normal power generation mode, storm warning mode, and extreme shutdown mode;

[0070] Based on the multi-state dynamic matching mechanism of the intelligent classification model, the environmental state is quickly and accurately identified, avoiding the adaptability defects of manually setting static thresholds. By coupling multi-parameter boundary dynamic correlation rules, the identification ability of extreme weather precursors is enhanced to provide a key decision window for control strategy switching and reduce passive response under sudden operating conditions;

[0071] A Markov state transition probability matrix is constructed based on historical operating data, and the environmental parameter change rate characteristics at each state transition are recorded;

[0072] By quantifying the historical state migration rules and environmental parameter change characteristics, a dynamic probability prediction framework is constructed to enhance the prediction ability of the gradual environmental deterioration process. Combined with parameter change trend and state transition correlation analysis, more accurate maintenance cycle planning and spare parts scheduling strategies are supported to optimize equipment life cycle management;

[0073] Based on the preset control parameters of the wind turbine generator unit under each environmental operating state, the adaptation degree of the wind turbine generator unit preset control parameters under each environmental operating state to other environmental operating states is determined;

[0074] Through cross-state control parameter adaptability analysis, the compatibility of the current control strategy to future potential operating conditions is evaluated to reduce mechanical impact and power fluctuations during strategy switching. Under the premise of ensuring core safety indicators, the adaptation scheme that takes into account power generation efficiency and equipment stability is preferred to improve the overall reliability of the system;

[0075] Based on the environmental parameter change rate characteristics at each state transition, the environmental parameter changes in the future period are predicted, and the hybrid prediction results are generated by combining with the state transition probability matrix;

[0076] Combining the dynamic characteristics of real-time environment and long-term statistical rules, a hybrid prediction model is constructed to effectively overcome the blind area of sudden disturbance for traditional prediction methods. By enhancing the time coverage range and prediction robustness, proactive decision-making basis is provided for active control such as unit yaw adjustment and power curve optimization, thereby improving the stability of the power generation process;

[0077] According to the comprehensive judgment of the prediction result and the adaptation degree evaluation, one of the control strategies of maintaining operation, parameter adjustment or emergency shutdown is triggered.

[0078] Through the multi-objective optimization decision mechanism, safety and economic indicators are dynamically balanced to reduce unnecessary shutdown caused by environmental misjudgment. In the case of extreme working condition warning, the protection strategy is triggered first, and in the normal working condition, smooth parameter adjustment is realized based on the adaptation degree evaluation, so as to achieve the dual goals of power generation efficiency optimization and equipment risk prevention and control

[0079] Referring to Figure 2 , the specific training process of the classification model is as follows:

[0080] An environmental parameter clustering center vector is set for each environmental running state respectively;

[0081] Based on the historical control log of the offshore wind farm, a plurality of environmental parameter vectors are determined for each environmental running state adjustment, denoted as sample environmental parameter vectors;

[0082] Based on the plurality of sample environmental parameter vectors and the environmental parameter clustering center vector, the environmental parameter boundary value of each environmental running state is determined;

[0083] Based on the environmental parameter clustering center vector and the environmental parameter boundary value, the classification model of the environmental running state is determined.

[0084] Through the cooperative training mechanism of the clustering center vector and the historical sample data, an adaptive environmental state classification model is constructed to overcome the experience dependency problem of traditional manual setting of classification boundaries. Based on the dynamic correction of the parameter boundary by the clustering center, the recognition accuracy of the transition region between normal, warning and shutdown states is improved, effectively avoiding the misjudgment risk of complex coupled factors such as sudden change of salt mist concentration and wave resonance. At the same time, through the boundary learning driven by historical sample data, the representation ability of the classification model to the nonlinear characteristics of the offshore environment is enhanced, providing high confidence state discrimination basis for preventive maintenance and dynamic control strategy generation.

[0085] Referring to Figure 3 , a Markov state transition probability matrix is constructed based on historical operation data, and the environmental parameter change rate characteristics at each state transition are recorded. Specifically, the Markov state transition probability matrix is constructed based on the historical operation data, and the environmental parameter change rate characteristics at each state transition are recorded. Specifically,

[0086] Based on the control properties of the wind turbine generator set, an analysis time is determined. The analysis time is determined by the control delay of the wind turbine generator set, and the maximum value of the time from the issuance of the control command to the completion of the parameter adjustment of the wind turbine generator set is taken as the analysis time.

[0087] Set a sample collection time, and set the time period within the sample collection time closest to the current time as the sample collection period;

[0088] Based on the analysis duration, the sample collection period is divided into several analysis periods;

[0089] Determine the environmental status for each analysis period based on the classification model;

[0090] Based on the environmental state of each analysis period, a Markov state transition probability matrix is ​​constructed. , is the Markov state transition probability from the i-th environmental operating state to the j-th environmental operating state, and K is the total number of preset environmental operating state types;

[0091] in, , is the total number of analysis periods in the i-th environmental state, is the number of adjacent analysis periods during which the environmental state changes from the i-th environmental state to the j-th environmental operating state;

[0092] The average values ​​of environmental parameters in each analysis period and the rate of change of environmental parameters over time in each analysis period are counted respectively, and the characteristics of the rate of change of environmental parameters during state transition are summarized.

[0093] By dynamically setting the analysis duration to match the unit control delay, the state transition probability calculation is ensured to be strictly synchronized with the actual response characteristics of the unit, thereby improving the temporal consistency between the prediction model and the execution action. Based on the division of real-time sample collection periods and the construction of the Markov state transition probability matrix, the correlation characteristics between the environmental state migration law and the parameter change rate are quantified, significantly enhancing the prediction reliability of short-term sudden disturbances (such as sudden changes in wind speed). At the same time, through statistical modeling of the environmental parameter change rate, the dynamic evolution trend during the state transition process is accurately captured, providing key prior information for the pre-adjustment of the unit control parameters, effectively reducing the risk of equipment damage caused by state switching delays, and improving the global stability of the system under complex sea conditions.

[0094] Determining the adaptability of the preset control parameters of the wind turbine generator set in each environmental operating state to other environmental operating states specifically includes:

[0095] Determine the performance index of the wind turbine under the preset control parameters of the wind turbine in each environment running state, the performance index at least including yaw delay, power fluctuation rate and blade stress peak value;

[0096] The calculation formula of the adaptation degree is:

[0097]

[0098] In the formula, is the adaptation degree of the preset control parameters of the wind turbine in the kth environment running state to the hth environment running state, is the total number of performance indexes, is the nth performance index corresponding to the preset control parameters of the wind turbine in the kth environment running state, is the nth performance index corresponding to the preset control parameters of the wind turbine in the hth environment running state, is the weight of the nth performance index.

[0099] Through the multi-dimensional performance index cross-state adaptation degree quantitative evaluation, a control parameter compatibility analysis framework is constructed, and the limitation of traditional single state parameter fixed configuration is broken through. Based on the weighted calculation model of performance difference, the short board of the current control strategy in the potential future working condition is dynamically identified. For example, the stress exceeding standard risk of extreme shutdown parameters in the early warning state, the adaptation scheme with the smallest mechanical impact and controllable power fluctuation is preferentially selected. Through the flexible adjustment of the weight coefficient, the priority of safety and economy indexes is adjusted, such as focusing on stress suppression in storm mode, realizing intelligent optimization of control strategy in different environmental evolution stages, and improving the robustness of the unit in all working conditions

[0100] Referring to Figure 4 , the future period environmental parameter change is predicted based on the environmental parameter change rate characteristics at each state transition, which specifically includes:

[0101] Based on the real-time parameter value of the real-time collected environmental parameter and the real-time change rate of the environmental parameter, and the average value of the environmental parameter in each analysis period and the change rate of the environmental parameter with time, the vector distance is solved, and the similarity distance index of the real-time environment and the environment in each analysis period is obtained;

[0102] Statistical next analysis period of each analysis length of the environmental running state;

[0103] Based on the similarity distance index of the real-time environment and the environment in all analysis periods and the transfer fitting probability of each environment running state in the future analysis length, the next analysis period of each analysis period is calculated;

[0104] Specifically, the calculation formula is:

[0105]

[0106] wherein, is the transition fitting probability of the i-th environmental operating state in the future analysis time length, is the analysis period set of the i-th environmental operating state in the next analysis period, c is the c-th element in the set, E is the set of all analysis periods, o is the o-th element in E, is the maximum value of the similarity distance index between the real-time environment and the environment in all analysis periods, is the minimum value of the similarity distance index between the real-time environment and the environment in all analysis periods, is the similarity distance index between the real-time environment and the environment of c.

[0107] By the dynamic similarity matching mechanism of real-time environment parameters and historical periods, a state transition fitting probability calculation model based on data driving is constructed, effectively solving the problem of insufficient sensitivity of traditional static prediction methods to real-time changes. The similarity between real-time environment and historical scene is quantified by vector distance, combined with the statistical law of the next state, and the influence trend of sudden disturbance on state transition is accurately captured. By normalizing the similarity distance weight distribution, high correlation historical samples are dynamically selected, and the adaptability of the prediction result to nonlinear environmental evolution is enhanced, providing high credibility basis for unit switching control strategy in advance, significantly reducing the risk of power oscillation or protection lag caused by prediction deviation.

[0108] Referring to Figure 5 shown, the specific steps of generating a hybrid prediction result by fusing the state transition probability matrix are as follows:

[0109] determining the environmental operating state at the current time, based on the environmental operating state at the current time, determining the Markov state transition probability of transitioning to each environmental operating state next from the Markov state transition probability matrix;

[0110] respectively setting the mixing weights of the transition fitting probability and the Markov state transition probability, and performing weighted summation on the transition fitting probability and the Markov state transition probability to obtain a hybrid prediction result.

[0111] Specifically, in some embodiments, the transition fitting probability and the Markov state transition probability are taken as 0.7 and 0.3 respectively, and the hybrid prediction result is 0.7 x transition fitting probability + 0.3 x Markov state transition probability.

[0112] By fusing the long-term statistical law of Markov model and the short-term dynamic characteristics of real-time fitting probability, a hybrid prediction mechanism is constructed to effectively balance the dual influence of historical experience and real-time environmental evolution. Based on a dynamic weight distribution strategy, the decision weight of real-time fitting probability is strengthened in the sudden disturbance scene to capture rapid changes, and the long-term trend prediction of Markov probability is focused on in the steady state condition to enhance the adaptability of the prediction result to complex environmental patterns. Through the complementary optimization of the dual models, the false negative risk in extreme conditions and the false positive risk in steady state are significantly reduced, providing a prediction basis for the unit control strategy that takes into account timeliness and stability, and realizing the upgrade of the decision-making mode from passive response to active defense.

[0113] Specifically, according to the comprehensive judgment of the prediction result and the adaptation degree evaluation, triggering one of the control strategies of maintaining operation, parameter adjustment or emergency shutdown specifically includes:

[0114] If the prediction probability of the extreme shutdown mode in the hybrid prediction result exceeds 0.5, emergency shutdown is triggered;

[0115] Otherwise, the prediction probability of each environmental operating state in the hybrid prediction result is used as the weight to weight the sum of the preset control parameters of the wind turbine generator set and the environmental adaptation degree in each environmental operating state, to obtain the prediction adaptation degree of the preset control parameters of the wind turbine generator set and the environment in each environmental operating state;

[0116] If the environmental operating state corresponding to the maximum prediction adaptation degree is the same as the current environmental operating state, maintaining operation is triggered;

[0117] If the environmental operating state corresponding to the maximum prediction adaptation degree is not the same as the current environmental operating state, parameter adjustment is triggered, and the control parameters of the offshore wind farm are adjusted to the preset control parameters of the wind turbine generator set in the environmental operating state corresponding to the maximum prediction adaptation degree.

[0118] Specifically, the calculation formula of the prediction adaptation degree is:

[0119]

[0120] Wherein, is the prediction adaptation degree of the preset control parameters of the wind turbine generator set and the environment in the i-th environmental operating state, is the prediction probability of the k-th environmental operating state in the hybrid prediction result, is the adaptation degree of the preset control parameters of the wind turbine generator set in the i-th environmental operating state to the k-th environmental operating state, and K is the total number of preset environmental operating states.

[0121] The prediction probability and the adaptation degree evaluation are fused through a multi-level decision mechanism to construct a control strategy generation framework balancing safety and economy dynamically. An emergency shutdown is triggered based on an extreme working condition probability threshold to ensure rigid constraints of core safety indicators of the equipment; in a normal scenario, a weighted fusion calculation of the prediction adaptation degree is performed to comprehensively consider future multi-state evolution trends and control parameter compatibility to preferentially select a globally optimal solution. This mechanism can avoid over-conservative or risky decisions caused by a single prediction model and can avoid the lag of dynamic environment response caused by static adaptation degree evaluation, thereby achieving a coordinated improvement of power generation efficiency optimization and risk prevention and control. Meanwhile, the robustness of the control strategy to uncertain environment evolution is enhanced through probabilistic mapping of the adaptation degree weight, and energy loss and mechanical wear caused by unnecessary mode switching are reduced.

[0122] Further, based on the same inventive concept as the above method, the present scheme also proposes a sea wind power plant control system, comprising:

[0123] A sensor array is arranged in the sensor array of the wind turbine generator set to collect wind speed, significant wave height and salt mist concentration data in real time.

[0124] A data processing module is configured to integrate the collected real-time data into a three-dimensional environmental parameter vector.

[0125] A state classification module is configured to match the three-dimensional environmental parameter vector to a plurality of preset environmental operating states including a normal power generation mode, a storm warning mode and an extreme shutdown mode by using a pre-trained classification model.

[0126] A Markov analysis module is configured to construct a state transition probability matrix based on historical operating data and store the environmental parameter change rate characteristics at each state transition.

[0127] An adaptation degree evaluation module is configured to calculate the control adaptation degree of each environmental operating state to other states according to preset control parameters of each environmental operating state.

[0128] An environmental prediction module is configured to generate a hybrid prediction result by combining the real-time environmental parameter change rate and the state transition probability matrix.

[0129] A control strategy execution module is configured to trigger one of a maintenance operation, a parameter adjustment or an emergency shutdown based on the hybrid prediction result and the adaptation degree evaluation result.

[0130] Further, the method according to the embodiments of the present application can also be implemented by means of Figure 6 the architecture of the electronic device as shown. As Figure 6As shown, the electronic device 500 can include a bus 501, one or more CPUs 502, a read-only memory (ROM) 503, a random access memory (RAM) 504, a communication port 505 connected to a network, an input / output component 506, a hard disk 507, etc. The storage device in the electronic device 500, such as the ROM 503 or the hard disk 507, can store the offshore wind farm control method provided in the present application. The electronic device 500 can also include a user interface 508. Of course, Figure 6 The architecture shown is only exemplary, and in implementing different devices, some of the components shown in the electronic device can be omitted Figure 6 The architecture shown is only exemplary, and in implementing different devices, some of the components shown in the electronic device can be omitted

[0131] Figure 7 is a computer readable storage medium structure provided by an embodiment of the present application. As shown in the figure, Figure 7 As shown, the computer readable storage medium 600 according to an embodiment of the present application. The computer readable storage medium 600 stores computer readable instructions. When the computer readable instructions are run by the processor, the offshore wind farm control method according to the embodiment of the present application described with reference to the above figures can be executed. The storage medium 600 includes, but is not limited to, for example, volatile memory and / or non-volatile memory. The volatile memory may, for example, include random access memory (RAM) and cache memory, etc. The non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0132] In summary, the advantages of the present application are as follows: the present application fuses multi-source environmental parameter real-time perception and intelligent state classification, combines dynamic Markov state transition prediction and cross-state adaptation evaluation, and significantly improves the self-adaptability of offshore wind farms to complex sea conditions. The multi-dimensional environmental parameter collaborative analysis and hybrid prediction mechanism effectively improves the extreme working condition warning accuracy, avoids the misjudgment or response lag caused by traditional static threshold; by predicting the environmental evolution trend and the adaptability of cross-state control parameters, the generator set operation strategy is dynamically optimized, the unnecessary downtime and frequent parameter adjustment are reduced, the equipment safety is ensured, and the power generation efficiency is improved. In addition, the fusion prediction model based on historical data and real-time change rate enhances the robustness of the system to sudden environmental disturbances, realizes the reduction of power generation loss and the reduction of equipment maintenance cost.

[0133] The basic principles, main features and advantages of the present application are shown and described above. Those skilled in the art should understand that the present application is not limited to the above embodiments, and the above embodiments and descriptions in the specification are only the principles of the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection claimed by the present application is defined by the appended claims and their equivalents.

Claims

1. A method of controlling an offshore wind farm, characterized by, The application relates to a wind turbine control method and system. Real-time wind speed, effective wave height and salt mist concentration data are collected through a sensor array arranged on a wind turbine generator set to form a three-dimensional environmental parameter vector; Real-time environmental parameters are input into a pre-trained classification model and matched to a plurality of preset environmental operating states, including at least a normal power generation mode, a storm warning mode and an extreme shutdown mode; A Markov state transition probability matrix is constructed based on historical operating data, and the environmental parameter change rate characteristics at state conversion are recorded; The adaptation degree of the preset control parameters of the wind turbine generator set in each environmental operating state to other environmental operating states is determined based on the preset control parameters of the wind turbine generator set in each environmental operating state; The future environmental parameter change is predicted based on the environmental parameter change rate characteristics at state conversion, and a hybrid prediction result is generated by fusing the state transition probability matrix; According to the comprehensive judgment of the prediction result and the adaptation degree evaluation, one of the control strategies of maintaining operation, parameter adjustment or emergency shutdown is triggered; The specific training process of the classification model is as follows: An environmental parameter clustering center vector is set for each environmental operating state; Based on the historical control log of the offshore wind farm, a plurality of environmental parameter vectors at the state adjustment time of each environmental operating state are determined, which are denoted as sample environmental parameter vectors; Based on the plurality of sample environmental parameter vectors and the environmental parameter clustering center vector, the environmental parameter boundary value of each environmental operating state is determined; Based on the environmental parameter clustering center vector and the environmental parameter boundary value, the classification model of the environmental operating state is determined.

2. A method of controlling an offshore wind farm according to claim 1, characterized in that, The specific process of constructing the Markov state transition probability matrix based on the historical operating data and recording the environmental parameter change rate characteristics at state conversion is as follows: Based on the control attributes of the wind turbine generator set, an analysis duration is determined; A sample collection duration is set, and the time period closest to the current time within the sample collection duration is set as a sample collection period; Based on the analysis duration, the sample collection period is divided into a plurality of analysis periods; The environmental state of each analysis period is determined based on the classification model; constructing a Markov state transition probability matrix based on the environment state of each analysis period, , Markov state transition probability from the i-th environment running state to the j-th environment running state, and K is a preset total number of environment running states. wherein, , is the total number of analysis periods in the i-th environmental state, is the number of adjacent analysis periods in which the environmental state changes from the i-th environmental state to the j-th environmental operating state. The average value of the environmental parameter in each analysis period and the change rate of the environmental parameter with time in each analysis period are respectively counted, and the environmental parameter change rate characteristics at state conversion are obtained by summarizing.

3. A method of controlling an offshore wind farm according to claim 2, characterized in that, The specific process of determining the adaptation degree of the preset control parameters of the wind turbine generator set in each environmental operating state to other environmental operating states is as follows: The performance indicators of the wind turbine generator set under the preset control parameters of the wind turbine generator set in each environmental operating state are determined, and the performance indicators at least include yaw delay, power fluctuation rate and blade stress peak value; The calculation formula of the adaptation degree is as follows: In the formula, is the degree of adaptation of the preset control parameter of the wind turbine under the kth environmental operating state to the hth environmental operating state, is the total number of performance indicators, is the nth performance indicator corresponding to the preset control parameter of the wind turbine under the kth environmental operating state, is the nth performance indicator corresponding to the preset control parameter of the wind turbine under the hth environmental operating state, is the side weight of the nth performance indicator.

4. A method of controlling an offshore wind farm according to claim 3, characterized in that, The specific process of predicting the future environmental parameter change based on the environmental parameter change rate characteristics at state conversion is as follows: Based on the real-time parameter value of the real-time collected environmental parameter and the real-time change rate of the environmental parameter, the average value of the environmental parameter in each analysis period and the change rate of the environmental parameter with time are solved to obtain the environmental similarity distance index of the real-time environment and each analysis period; The environmental operating state in the next analysis period of each analysis duration is counted. The transition fitting probability of each environment running state in the future analysis period is calculated based on the real-time environment, the similar distance index of the environment in all analysis periods, and the environment running state in the next analysis period of each analysis period. Specifically, the calculation formula is: wherein, is the transition fitted probability of the i-th environmental operating state within the future analysis time length, is the analysis time period set of the i-th environmental operating state within the next analysis time period, c is the c-th element in, E is the set of all analysis time periods, o is the o-th element in E, is the maximum value of the similarity distance indicator between the real-time environment and the environment within all analysis time periods, is the minimum value of the similarity distance indicator between the real-time environment and the environment within all analysis time periods, is the similarity distance indicator between the real-time environment and the environment of c.

5. A method of controlling an offshore wind farm according to claim 4, characterized in that, The specific steps of generating the mixed prediction result by fusing the state transition probability matrix are: Determine the current environment running state, and determine the Markov state transition probability of the next transition to each environment running state from the Markov state transition probability matrix based on the current environment running state. The mixing weights of the transition fitting probability and the Markov state transition probability are set respectively, and the transition fitting probability and the Markov state transition probability are weighted and summed to obtain the mixed prediction result.

6. A method of controlling an offshore wind farm according to claim 5, characterized in that, The specific control strategy triggered by the comprehensive judgment of the prediction result and the adaptation degree evaluation includes: If the prediction probability of the extreme shutdown mode in the mixed prediction result exceeds 0.5, emergency shutdown is triggered; Otherwise, the prediction probability of each environment running state in the mixed prediction result is used as the weight to weight and sum the adaptation degree of the wind turbine group preset control parameter and the environment running state under each environment running state, to obtain the prediction adaptation degree of the wind turbine group preset control parameter and the environment under each environment running state. If the environment running state corresponding to the maximum prediction adaptation degree is the same as the current environment running state, maintain operation is triggered; If the environment running state corresponding to the maximum prediction adaptation degree is not the same as the current environment running state, parameter adjustment is triggered, and the control parameter of the offshore wind farm is adjusted to the wind turbine group preset control parameter under the environment running state corresponding to the maximum prediction adaptation degree.

7. A control system for an offshore wind farm, configured to implement the control method of the offshore wind farm according to any one of claims 1-6, comprising: a sensor array arranged on the wind turbine group, configured to collect wind speed, significant wave height and salt mist concentration data in real time; a data processing module configured to integrate the collected real-time data into a three-dimensional environmental parameter vector; a state classification module, which internally stores a pre-trained classification model, and matches the three-dimensional environmental parameter vector to a plurality of preset environment running states, including normal power generation mode, storm warning mode and extreme shutdown mode; a Markov analysis module, configured to construct a state transition probability matrix based on historical operation data, and store the environmental parameter change rate characteristics at each state transition; an adaptation degree evaluation module, configured to calculate the control adaptation degree of each environment running state to other states according to the preset control parameter under each environment running state; an environment prediction module, configured to generate a mixed prediction result by combining the real-time environmental parameter change rate and the state transition probability matrix; a control strategy execution module, configured to trigger one of the control instructions of maintaining operation, parameter adjustment or emergency shutdown according to the mixed prediction result and the adaptation degree evaluation result.

8. An electronic device, comprising: comprising: at least one processor; and a memory connected in communication with the at least one processor; wherein, The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the offshore wind farm control method of any of claims 1-6.

9. A computer-readable storage medium storing computer-readable instructions, the computer-readable instructions comprising: The computer readable instructions, when executed by the processor, implement the offshore wind farm control method of any of claims 1-6.

Images