Wave energy power generation efficiency optimization control method and device based on artificial intelligence

Abstract

The application discloses a wave energy power generation efficiency optimization control method and device based on artificial intelligence, S1. Collect marine environment data set; S2. Form the marine environment data set after preprocessing; S3. According to the preprocessed marine environment data set and the structure parameters of the wave energy power generation system, an optimization objective function is constructed; S4. The optimization objective function is globally searched by using the bitter fish optimization algorithm, the optimization population is initialized, and a plurality of candidate optimal parameter regions are determined; S5. The adaptive jellyfish search algorithm is used for local fine search on the region where the candidate optimal parameters are located, the key control parameters are updated and optimized, and the key control parameters for further improving the electric energy conversion efficiency are obtained; S6. The key control parameters are applied to the wave energy power generation system to form a dynamic closed-loop feedback control. The application ensures that the system is always in the best working state, thereby further improving the electric energy conversion efficiency and significantly reducing the energy loss.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wave energy generation, and in particular to a wave energy generation efficiency optimization control method and device based on artificial intelligence. BACKGROUND

[0002] With the increasing demand for renewable energy worldwide, wave energy has received widespread attention as a green and sustainable energy source. Current wave energy generation control strategies mostly use fixed parameters or empirical formulas to construct optimization objective functions and constraints, failing to fully consider the dynamic changes and nonlinear characteristics of wave environments. Static control methods have obvious limitations in practical applications: on the one hand, the system has difficulty in achieving timely and accurate dynamic response and parameter adjustment when encountering sudden sea conditions or extreme environments; on the other hand, traditional optimization algorithms are deficient in global search and local fine optimization, often falling into local optimal state, failing to fully explore and exploit the optimal operating parameters of wave energy generation systems, resulting in difficulty in achieving ideal levels of energy conversion efficiency and output power stability.

[0003] However, existing technologies generally ignore the systematic processing of marine environment monitoring data and the construction of intelligent decision-making capabilities. Most current wave energy generation systems rely on fixed control strategies or manually set parameters, lacking depth understanding and dynamic response capabilities for complex environmental factor data. In fact, the availability of wave energy is closely related to the dynamic changes of wave height, period, wind speed, and current speed. To achieve optimal control of system performance, an intelligent control method is needed that can receive and process multi-source environmental monitoring data in real time, analyze the influence mechanism of the environment on generation efficiency, and dynamically generate key control parameters to adapt to rapidly changing sea conditions. SUMMARY

[0004] One object of the present application is to provide a wave energy generation efficiency optimization control method and device based on artificial intelligence, which ensures that the system is always in the best working state, thereby further improving the energy conversion efficiency and significantly reducing the energy loss.

[0005] According to the wave energy generation efficiency optimization control method based on artificial intelligence, the method comprises the following steps:

[0006] S1. Collecting marine environment data sets;

[0007] S2. Preprocessing the collected marine environment data sets to form preprocessed marine environment data sets;

[0008] S3. Constructing an optimization objective function according to the preprocessed marine environment data sets and the structural parameters of the wave energy generation system, and determining the corresponding constraints;

[0009] S4. Global search of the optimization objective function is performed by using the bitter fish optimization algorithm, an optimization population is initialized, and rough search of the key control parameters of the wave energy generation system is performed according to fitness evaluation, so as to determine a plurality of candidate optimal parameter regions;

[0010] S5. Local fine search of the region where the candidate optimal parameters are located is performed by using the adaptive jellyfish search algorithm, the key control parameters are updated and optimized, and the key control parameters for further improving the power conversion efficiency are obtained;

[0011] S6. The key control parameters are applied to the wave energy generation system, and dynamic closed-loop feedback control is formed by combining the real-time monitored marine environment data and the operation state of the power generation system, so as to adjust the parameters of the wave energy generation system in real time.

[0012] Optionally, the S1 includes the following steps:

[0013] S11. The marine environment data set of the wave energy generation system is defined , and the marine environment data set includes effective wave height, average wave period, average wind speed, and average current speed:

[0014]

[0015] wherein, represents the effective wave height, represents the average wave period, represents the average wind speed, represents the average current speed;

[0016] S12. The time series set of environmental data is constructed according to the measurement requirements of the wave energy generation system, and the marine environment data set of the wave energy generation system at time is defined .

[0017] S13. The multi-point measurement method is adopted to collect the marine environment data set at different measurement points of the wave energy generation system, and the marine environment data set with spatial distribution is constructed in combination with the measurement points:

[0018]

[0019] wherein, , , , respectively represent the effective wave height, the average wave period, the average wind speed, and the average current speed at the measurement point .

[0020] Optionally, the S2 includes the following steps:

[0021] S21. Perform integrity check on the marine environment dataset, eliminate incomplete data records, and make each data record contain valid wave height, average wave period, average wind speed, and average current speed, to form a complete marine environment dataset;

[0022] S22. Perform outlier detection on the marine environment dataset, eliminate data points beyond the range, and form a marine environment dataset after outlier elimination;

[0023] S23. Perform data smoothing processing on the marine environment dataset after outlier elimination using a noise filtering method, to construct a filtered marine environment dataset;

[0024] S24. Perform normalization processing on the filtered marine environment dataset, to normalize all environmental parameters to the same numerical range, to form a normalized marine environment dataset ;

[0025] S25. Resample the normalized marine environment dataset according to the time step to ensure consistent time intervals of the data, to form a final preprocessed marine environment dataset:

[0026] .

[0027] Optionally, the S3 includes the following steps:

[0028] S31. Extract feature variables related to the operation of the wave energy power generation system from the preprocessed marine environment dataset to construct an environmental parameter vector , and based on the preprocessed marine environment dataset and the structural parameters of the wave energy power generation system, establish a mapping relationship between the environmental parameters and the key performance variables of the system:

[0029]

[0030] wherein, is a mapping function, represents a set of optimization variables of the wave energy power generation system, used to optimize the control of the electric energy conversion efficiency, output power stability, and energy loss:

[0031]

[0032] wherein, is the active power output of the wave energy power generation system, is the conversion efficiency of the wave energy power generation system, is the motion angle of the wave energy conversion device, ​the natural frequency of the wave energy device, the damping coefficient of the wave energy converter;

[0033] S32. According to the optimization objective of the wave energy generation system, combine the pre-processed ocean environment data set and the optimization variable set , establish the optimization objective function:

[0034]

[0035] wherein, is the weight coefficient used to balance the electric energy conversion efficiency, output power stability and energy loss in the optimization process, is the electric energy conversion efficiency, indicating the proportion of wave energy effectively converted into electric energy:

[0036]

[0037] wherein, is the wave input power, determined by the effective wave height and the average wave period:

[0038]

[0039] wherein, is the gravitational acceleration, is the average wave period;

[0040] is the output power stability, indicating the smoothness of the generated power:

[0041]

[0042] wherein, is the average output power of the wave energy generation system within the preset time, is the total number of time steps;

[0043] is the energy loss, indicating the energy loss in the system operation:

[0044]

[0045] wherein, is the resistance loss of the power generation system, is the generator current, is the mechanical energy loss, determined by the average wave period and the motion angle of the wave energy conversion device ;

[0046] S33. According to the operation constraints of the wave energy generation system, set the optimization variable set The constraint conditions are:

[0047]

[0048]

[0049] ;

[0050]

[0051]

[0052] wherein, Pmin and Pmax represent the minimum and maximum output power respectively, ηmin and ηmax represent the minimum and maximum conversion efficiency respectively, θmin and θmax represent the minimum and maximum motion angle respectively, ωmin and ωmax represent the minimum and maximum natural frequency respectively, ζmin and ζmax represent the minimum and maximum damping coefficient respectively;

[0053] Set the environmental adaptability constraint to ensure that the wave energy generation system remains stable in complex marine environments:

[0054]

[0055] wherein, Pmax represents the maximum power change rate allowed by the system;

[0056] S34. Based on the construction of the optimization objective function and the setting of the constraint conditions, the final optimization solving problem is formed:

[0057]

[0058] wherein, Ω represents all feasible solution spaces that meet the constraint conditions.

[0059] Optionally, the S4 comprises the following steps:

[0060] S41. Based on the optimization objective function and the optimization variable set Define the environmental adaptive guppy optimization population, set the parameter vector of the optimization individual, and adjust the population distribution according to the environmental dynamic characteristics of the wave energy generation system:

[0061]

[0062] wherein, N represents the size of the guppy optimization population, represents the optimization variable vector of the th individual:

[0063]

[0064] in, For the Active power output of each individual wave energy generation system, For the The conversion efficiency of individual wave energy power generation systems, For the The movement angle of each individual wave energy conversion device, For the The natural frequency of each individual wave energy device, For the The damping coefficient of each individual wave energy power generation device is calculated; the population is dynamically initialized and optimized based on the environmental characteristics of the wave energy power generation system, and the environmental impact factor is constructed based on the marine environmental data set. :

[0065]

[0066] Through environmental factors Adjust the initial population distribution to adapt the initial population to the optimization calculation under different sea conditions:

[0067]

[0068] in, for A random number between and are the minimum and maximum allowed values ​​of the optimization variables, respectively;

[0069] S42. Optimizing the population Each individual in , combined with the environmental adaptability of the wave energy power generation system, an environmental constraint adaptive fitness function is constructed:

[0070]

[0071] in, To adaptively adjust the weight of environmental constraints, reduce the weight when the sea conditions are stable and increase the weight when the sea conditions change drastically. is the environmental constraint penalty factor, which represents the fitness deviation of individuals in different marine environments:

[0072]

[0073] in, It is the optimal solution under historical circumstances. Number of samples stored for the environment;

[0074] S43. Based on the environment adaptive bitterfish optimization strategy, a dynamic step adjustment mechanism is introduced in the search process, and the optimization variables are updated:

[0075]

[0076] wherein, is the optimization variable set of the th individual in the th generation, is the current optimal individual, is the step factor, is a random number between [0, 1], is the adaptive disturbance amplitude;

[0077] S44. A dynamic aggregation strategy is adopted to enhance the optimization ability of bitterfish population. The dynamic aggregation strategy is introduced to improve the convergence speed of bitterfish population. The attraction between bitterfish individuals is defined , and the behavior of bitterfish population is simulated:

[0078]

[0079] wherein, avoiding zero denominator, the larger the value is, the more the individual tends to the individual ;

[0080] When updating the individual position, the aggregation behavior of the population is combined:

[0081]

[0082] wherein, controls the interaction strength between individuals;

[0083] S45. The S42-S44 process is iterated multiple times to finally determine the candidate optimal parameter region :

[0084]

[0085] wherein, is the fitness threshold, only the candidate optimal solution that meets the target performance is retained.

[0086] Optionally, the S5 comprises the following steps:

[0087] S51. According to the determined candidate optimal parameter region , a local search population of adaptive jellyfish search algorithm is constructed, and the parameter vector of the local search individual is set:

[0088]

[0089] wherein, denotes the key control parameter vector of the i-th individual in the candidate region;

[0090] S52. For each individual in the local search population , a local fine search is performed using an adaptive jellyfish search algorithm to update the individual parameters:

[0091]

[0092] wherein, denotes the parameter vector of the i-th individual in the j-th generation, is the parameter vector of the individual with the highest fitness in the local search population, is the local search step size factor, is the local perturbation weight, is the local perturbation vector generated according to the jellyfish movement mechanism, reflecting the exploration ability of the jellyfish in the local region; S53. According to the change of the fitness of the individual in the local search process, the step size factor is adaptively adjusted:

[0093]

[0094] wherein,

[0095] is the step size attenuation control coefficient, used to adjust the step size update rate; S54. The steps S52 to S53 are iterated until the local search reaches the preset convergence condition or the maximum iteration number

[0096] is satisfied, and the key control parameters for further improving the power conversion efficiency are finally obtained:

[0097]

[0098] wherein, denotes the optimal key control parameter of the i-th individual at the end of the local search. The application discloses a wave energy power generation efficiency optimization control device based on artificial intelligence.

[0099] The application discloses a wave energy power generation efficiency optimization control device based on artificial intelligence.

[0100] The application has the beneficial effects that:​​​​

[0101] (1) The application adopts a hybrid optimization strategy combining the bitter fish optimization algorithm and the adaptive jellyfish search algorithm. The traditional optimization method has the problems of low global search efficiency or local optimal disturbance. The application first uses the bitter fish optimization algorithm to perform global search on the entire parameter space, quickly locates the candidate area of key parameters by adaptively adjusting the optimization population distribution of the environment, and then uses the adaptive jellyfish search algorithm to perform local fine search on the candidate area, realizes accurate search and local refinement in the parameter space, and ensures the accuracy of the control parameters and the significant improvement of the system energy conversion efficiency.

[0102] (2) The application introduces an environmental adaptability factor to real-time correct the initialization and dynamic adjustment of the optimization population, so that the system can quickly respond and adjust the control parameters in the face of sudden sea state changes, thereby maintaining high output power stability and low energy loss in complex and unstable marine environments, overcoming the defects of traditional methods that rely heavily on fixed parameters and lack of dynamic adaptability.

[0103] (3) The application designs a closed-loop feedback control mechanism based on real-time marine environment data and power generation system operating state. During the optimization process, the system continuously collects the latest environmental and equipment data, dynamically updates the key control parameters, forms a feedback loop, realizes adaptive adjustment, and the closed-loop mechanism can effectively cope with the variable interference factors in the marine environment, ensuring that the system is always in the best working state, thereby further improving the electric energy conversion efficiency and significantly reducing the energy loss. BRIEF DESCRIPTION OF DRAWINGS

[0104] The accompanying drawings are included to provide a further understanding of the application, and constitute a part of the specification, together with the embodiments of the application, to explain the application, and do not constitute a limitation on the application. In the drawings:

[0105] Figure 1 A flowchart of a wave energy power generation efficiency optimization control method and device based on artificial intelligence is proposed for the application. DETAILED DESCRIPTION

[0106] The application will now be described in further detail with reference to the accompanying drawings. These drawings are all simplified schematic diagrams, and only illustrate the basic structure of the application in a schematic manner, and therefore only show the components related to the application.

[0107] Reference Figure 1 A wave energy power generation efficiency optimization control method based on artificial intelligence includes the following steps:

[0108] S1. Collect marine environment data sets;

[0109] Specifically, a marine environment data set in a target sea area is collected, including wave data, tidal change data, wind speed data and current velocity data; and an optimization model for cluster arrangement of wave energy equipment is constructed based on the collected data, the optimization model taking maximizing unit area energy capture capacity, minimizing interference between equipment, minimizing total anchor system tension and drift risk as optimization objectives, and a multi-objective optimization expression is constructed.

[0110] S2. The collected marine environment data set is preprocessed to form a preprocessed marine environment data set;

[0111] S3. An optimization objective function is constructed according to the preprocessed marine environment data set and the structural parameters of the wave energy power generation system, and corresponding constraint conditions are determined;

[0112] S4. The optimization objective function is globally searched by using the stingray optimization algorithm, the optimization population is initialized, and the key control parameters of the wave energy power generation system are roughly searched according to the fitness evaluation, to determine several candidate optimal parameter regions;

[0113] S5. The self-adaptive jellyfish search algorithm is used to perform local fine search on the region where the candidate optimal parameters are located, to update and optimize the key control parameters, and to obtain the key control parameters for further improving the electric energy conversion efficiency;

[0114] S6. The key control parameters are applied to the wave energy power generation system, and combined with the real-time monitored marine environment data and the operation state of the power generation system, a dynamic closed-loop feedback control is formed, to adjust the wave energy power generation system parameters in real time, including the control device's response frequency and attitude to waves; the damping of the power generation module, the electromagnetic load characteristics; the stable slope and fluctuation range of the output power; and the maintenance of the safe operation state of the device in sudden sea conditions. Through the dynamic real-time adjustment of the wave energy power generation system parameters, the system can continuously maintain the optimal working state in complex and rapidly changing marine environments, and realize the dual improvement of energy conversion efficiency and stability.

[0115] In the system operation process in the embodiment, the marine environment data acquisition module records the key elements of significant wave height, period and average wind speed at a high frequency (sampling once every 10 seconds), but the parameter optimization calculation is not processed independently for each sampling point. Instead, to ensure the representativeness and statistical stability of the environmental data, the system sets a dynamic sliding time window ΔT, usually 2-5 minutes, to construct a short-term feature sequence of the marine environment. Within this time window, the system will process multiple groups of data for each measurement point as follows:

[0116] Short-term mean and variance calculation: used to measure the average state and volatility of elements such as significant wave height and period;

[0117] Trend analysis: determine whether there is a sustained upward or sudden trend in key environmental variables;

[0118] Event trigger mechanism: sudden increase in average wind speed beyond dynamic threshold, rapid shortening of average wave period, will trigger immediate optimization response.

[0119] Based on the above data processing flow, the shortest cycle of system optimization response is 1 minute, and the longest is not more than 5 minutes, to ensure:

[0120] On the one hand, it can respond quickly to sudden sea conditions, and does not cause energy loss or system instability due to long time delay; on the other hand, it avoids misjudgment or frequent parameter adjustment when the environmental fluctuation has not yet formed a trend, ensuring the stability of system operation and the convergence of controller.

[0121] In specific implementation, the present application also introduces a response delay scheduler module, which is based on the execution timestamp of the previous round of control parameter update , combined with the current environmental data change rate, dynamically adjusts the next round of optimization trigger time , to ensure reasonable response frequency. For example:

[0122] If the current sea condition is stable, the response interval is automatically extended to 5 minutes by the delay scheduler;

[0123] If three consecutive measurement points are detected to have a synchronous fluctuation trend, the response period is actively shortened to 1 minute.

[0124] In addition, in order to avoid excessive frequent parameter update leading to mechanical fatigue of equipment, the present application sets a minimum control parameter stable period , within which period, unless the sudden sea condition exceeds the system warning threshold, the control instruction is not actively updated.

[0125] In summary, in the real-time adjustment process of the embodiment, the sliding time window mechanism, event trigger mechanism and response delay scheduling strategy are introduced, which not only ensures the timeliness of control response, but also fully considers the statistical characteristics of marine environmental data and the engineering feasibility of system operation, thereby realizing the organic unification between dynamic closed loop and stable parameter adjustment.

[0126] In this embodiment, the monitoring device of the wave energy power generation system is usually integrated with the wave energy device body, but in special environments (such as the distance between the equipment is too large) a split design is used to increase the data volume.

[0127] In this embodiment, S1 includes the following steps:

[0128] S11. Define the marine environmental data set of the wave energy power generation system , the marine environment data set includes significant wave height, mean wave period, mean wind speed, mean current speed:

[0129]

[0130] wherein, represents the significant wave height, represents the mean wave period, represents the mean wind speed, represents the mean current speed;

[0131] S12. Constructing a time series set of environment data according to the measurement requirement of the wave energy power generation system, defining the marine environment data set of the wave energy power generation system at time ; ;

[0132] S13. Collecting marine environment data set at different measurement points of the wave energy power generation system by using multi-point measurement method, combining the measurement points to construct marine environment data set with spatial distribution:

[0133]

[0134] wherein, , , , respectively represent the significant wave height, mean wave period, mean wind speed, mean current speed at the measurement point .

[0135] In the embodiment, S2 includes the following steps:

[0136] S21. Performing integrity check on the marine environment data set , eliminating incomplete data records, so that each data contains significant wave height, mean wave period, mean wind speed, mean current speed, forming complete marine environment data set;

[0137] S22. Performing outlier detection on the marine environment data set, eliminating data points out of range, forming marine environment data set after outlier elimination;

[0138] S23. Performing data smoothing processing on the marine environment data set after outlier elimination by using noise filtering method, constructing filtered marine environment data set;

[0139] S24. Performing normalization processing on the filtered marine environment data set, normalizing all environment parameters to the same numerical range, forming normalized marine environment data set ;

[0140] S25. Constructing time series set of environment data according to time step The normalized marine environment dataset Resampling is performed to ensure the consistency of the time interval of the data, forming the final pre-processed marine environment dataset:

[0141] .

[0142] In this embodiment, after forming the final pre-processed marine environment dataset, a machine learning-based environmental parameter mapping model is constructed according to the pre-processed marine environment dataset and the wave energy power generation system structure parameters , the environmental parameter vector is constructed, and the mapping relationship between the environmental parameter vector and the system key performance variable is established: The mapping model is optimized using a machine learning method: a neural network or regression analysis is selected to train the mapping relationship. A hyperparameter optimization method is used to dynamically adjust the model parameters to improve the prediction accuracy: wherein is a set of model hyperparameters, is a model loss function. Combined with historical operation data, an incremental learning mechanism is used to continuously optimize the prediction accuracy: wherein is the learning rate, is the model parameter gradient.

[0143] In this embodiment, S3 includes the following steps:

[0144] S31. According to the pre-processed marine environment dataset extract the feature variables related to the operation of the wave energy power generation system, construct the environmental parameter vector , based on the pre-processed marine environment dataset and the structure parameters of the wave energy power generation system, establish the mapping relationship between the environmental parameter vector and the system key performance variable:

[0145]

[0146] wherein is a mapping function, represents a set of optimization variables of the wave energy power generation system, which is used to optimize the control of the power conversion efficiency, output power stability and energy loss:

[0147]

[0148] wherein is the active power output of the wave energy power generation system, is the conversion efficiency of the wave energy power generation system, is the motion angle of the wave energy conversion device, the natural frequency of the wave energy device, the damping coefficient of the wave energy power generation device;

[0149] S31. Introducing a machine learning model for feature mapping according to the machine learning model , mapping the environmental parameters to the optimization variables , ensuring the adaptability of the objective function construction: Optimization objective function dynamic updating mechanism: combined with the running data, adjust the weight of the objective function, make it conform to the dynamic characteristics of the wave energy power generation system: where, is the adjustment coefficient, is the weight update amount calculated based on historical data.

[0150] S32. According to the optimization objective of the wave energy power generation system, combined with the preprocessed marine environment data set and the optimization variable set , the optimization objective function is established:

[0151]

[0152] where, is the weight coefficient used to balance the power conversion efficiency, output power stability and energy loss in the optimization process, is the power conversion efficiency, which represents the proportion of wave energy effectively converted into electric energy:

[0153]

[0154] where, is the wave input power, determined by the effective wave height and the average wave period:

[0155]

[0156] where, is the acceleration of gravity, is the average wave period;

[0157] is the output power stability, which represents the smoothness of the power generation power:

[0158]

[0159] where, is the average output power of the wave energy power generation system within a preset time, is the total number of steps;

[0160] is the energy loss, which represents the energy loss in the system operation:

[0161]

[0162] wherein, R is the electrical resistance loss of the power generation system, I is the generator current, M is the mechanical energy loss, which depends on the average wave period and the motion angle of the wave energy conversion device ;

[0163] S33. Set the constraint conditions of the optimization variable set according to the operation constraints of the wave energy power generation system:

[0164]

[0165]

[0166] ;

[0167]

[0168]

[0169] wherein, Pmin and Pmax represent the minimum and maximum output power, respectively, ηmin and ηmax represent the minimum and maximum conversion efficiency, respectively, θmin and θmax represent the minimum and maximum motion angle, respectively, ωmin and ωmax represent the minimum and maximum natural frequency, respectively, ζmin and ζmax represent the minimum and maximum damping coefficient, respectively;

[0170] Set the environmental adaptability constraints to ensure that the wave energy power generation system remains stable in complex marine environments:

[0171]

[0172] wherein, ΔPmax represents the maximum power change rate allowed by the system;

[0173] S34. Based on the construction of the optimization objective function and the setting of the constraint conditions, the final optimization solving problem is formed:

[0174]

[0175] wherein, Ω represents the feasible solution space that meets all the constraint conditions.

[0176] In this embodiment, S4 includes the following steps:

[0177] S41. Based on the optimization objective function and the optimization variable set Define the environment adaptive grouper optimization population, set the parameter vector of the optimization individual, and adjust the population distribution according to the dynamic characteristics of the wave energy generation system environment:

[0178]

[0179] Wherein, is the size of the grouper optimization population, represents the optimization variable vector of the th individual:

[0180]

[0181] Wherein, is the active power output of the wave energy generation system of the th individual, is the conversion efficiency of the wave energy generation system of the th individual, is the motion angle of the wave energy conversion device of the th individual, is the natural frequency of the wave energy device of the th individual, is the damping coefficient of the wave energy generation device of the th individual; Combine the environmental characteristics of the wave energy generation system to dynamically initialize the optimization population, and construct the environmental influence factor according to the marine environment data set:

[0182]

[0183] Adjust the initial population distribution through the environmental influence factor , so that the initial population adapts to the optimization calculation under different sea conditions:

[0184]

[0185] Wherein, is a random number between and and are the minimum and maximum allowed values of the optimization variable respectively;

[0186] S42. For each individual in the optimization population , construct the environment constraint adaptive fitness function according to the environmental adaptability of the wave energy generation system:

[0187]

[0188] Wherein, To adaptively adjust the environmental constraint weight, reduce the weight when the sea condition is stable, and increase the weight when the sea condition changes dramatically, The environmental constraint penalty factor represents the fitness deviation of individuals in different marine environments:

[0189]

[0190] Wherein, The optimal solution in the historical environment, The number of environmental storage samples;

[0191] S43. Based on the adaptive bitterfish optimization strategy, a dynamic step adjustment mechanism is introduced in the search process to update the optimization variables:

[0192]

[0193] Wherein, The optimization variable set of the th individual in the th generation, The current optimal individual, The step factor, A random number between 0 and 1, The adaptive disturbance amplitude;

[0194] S44. Dynamic aggregation strategy is adopted to enhance the optimization ability of bitterfish population, dynamic aggregation strategy is introduced to improve the convergence speed of bitterfish population, and the attraction between bitterfish individuals is defined , simulate the behavior of bitterfish population:

[0195]

[0196] Wherein, Avoiding zero denominator, The larger the value, the more the individual tends to the individual ;

[0197] When updating the individual position, the group aggregation behavior is combined:

[0198]

[0199] Wherein, Control the strength of interaction between individuals;

[0200] S45. Multiple iterations are performed on the processes of S42-S44 to finally determine the candidate optimal parameter region :

[0201]

[0202] Wherein, is the fitness threshold, and only the candidate optimal solutions that meet the target performance are retained.

[0203] In this embodiment, S5 includes the following steps:

[0204] S51. Based on the determined candidate optimal parameter region , construct the local search population of the adaptive jellyfish search algorithm and set the parameter vector of the local search individual:

[0205]

[0206] in, Indicates the candidate area The key control parameter vector of each individual;

[0207] S52. For local search population Each individual in , use the adaptive jellyfish search algorithm to perform local fine search and update individual parameters:

[0208]

[0209] in, Indicates the Individuals in The parameter vector of the generation, is the individual parameter vector with the highest current fitness in the local search population, is the local search step factor, is the local perturbation weight, The local disturbance vector generated by the jellyfish movement mechanism reflects the jellyfish's exploration ability in the local area;

[0210] S53. According to the change of individual fitness during the local search process, the step factor Make adaptive adjustments:

[0211]

[0212] in, is the step size attenuation control coefficient, which is used to adjust the step size update rate;

[0213] S54. Iterate steps S52 to S53 until the local search reaches the preset convergence condition or the maximum number of iterations are satisfied, and finally the key control parameters for further improving the power conversion efficiency are obtained:

[0214]

[0215] in, represents the optimal key control parameter of the individual at the time of termination of the local search.

[0216] In this embodiment, an automatic control system is introduced to adjust the parameters, and a PID controller or model predictive control is used to optimize the key control parameters of the wave energy generation system.

[0217] PID control: wherein, is the error between the desired value and the current value, is the PID parameter.

[0218] MPC control: wherein, is the prediction horizon, is the target parameter, is the control input.

[0219] A safety threshold for parameter adjustment is set to set a safe operating range to prevent excessive parameter adjustment from affecting equipment stability: If the threshold condition is violated, the system triggers a safety protection mechanism to adjust the parameters to restore them to the safe range.

[0220] New optimization parameters for the power generation device are added to expand the optimization variables , including the distributed arrangement of the wave energy generation device and the adjustment of electromagnetic parameters

[0221] The equipment distributed arrangement optimization uses the stingray optimization + jellyfish search to optimize the distribution position of the power generation device in the ocean wherein, is the energy loss between the power generation devices, is the equipment arrangement parameter.

[0222] A wave energy generation efficiency optimization control device based on artificial intelligence includes a processor, a memory, and a wave energy generation efficiency optimization control program stored in the memory and executable by the processor, wherein when the wave energy generation efficiency optimization control program is executed by the processor, a wave energy generation efficiency optimization control method based on artificial intelligence is implemented.

[0223] Embodiment: In this test scenario, the wave energy generation system adopts a multi-point deployment strategy, and a total of 3 sets of power generation systems are deployed, located in three sub-areas around the main control platform P0 (coordinates 37.5003°N, 120.3003°E): A area, B area and C area, which form a triangular deployment network with a radius of about 800 meters. Each set of subsystem is equipped with a set of wave energy conversion devices, monitoring sensor modules and edge computing units for local data acquisition and preprocessing.

[0224] ​Point A is located about 100 meters west of the power generation device in Area A, mainly collecting local wave characteristics of the sea area where the subsystem is located to guide the adjustment of the A area power generation equipment parameters by the P0 central control system.

[0225] Point B is located 70 meters east-south of the power generation device in Area B, providing independent environmental input for Area B and serving as a boundary point to share data with Area C.

[0226] Point C is located near the northeast edge of Area C, about 780 meters from the main control platform, mainly serving as a remote sudden sea condition monitoring point to detect overall sea condition trends and provide rapid warning.

[0227] At 2:30 am on July 15, 2023, the monitoring system of a certain wave energy power station in a certain province received the latest marine environmental data from measurement point A (latitude and longitude coordinates 37.5001°N, 120.3005°E), which showed that the current effective wave height was 1.8 meters, the average wave period was 7.2 seconds, the average wind speed was 5.8 meters / second, the average current speed was 1.2 meters / second, at this time, the real-time power output of the power generation system was 470kW, the energy conversion efficiency was only 41.3%, and the power fluctuation was large, the maximum fluctuation range reached ±55kW.

[0228] The control strategy of the system adopts a central coordination + local autonomy architecture, when the data of measurement points A and B changes little and the equipment runs stably, the local edge nodes can independently complete parameter adjustment; when multiple measurement points change dramatically at the same time, the main control platform P0 will uniformly dispatch and coordinate the control strategies of the three areas to achieve global rapid response.

[0229] At the same time, measurement point B (about 700 meters away from point A) recorded different sea condition data: effective wave height 2.0 meters, average wave period 7.5 seconds, average wind speed 6.1 meters / second, average current speed 1.4 meters / second, while the power generation system still runs in the traditional fixed parameter mode and fails to adjust to the environmental data of measurement point B, resulting in further fluctuations in equipment power output, and at 2:45 am, the system power output dropped to 430kW, and the cumulative energy loss in 15 minutes reached 21kWh.

[0230] In order to verify the optimization control method of the application, the system starts the optimization calculation process based on the bitter fish optimization algorithm and the adaptive jellyfish search algorithm at 2:50. First, the system initializes the optimization population according to the real-time collected marine environment data, and performs global search. At 2:53, five candidate optimal parameter regions are determined, and the local fine search stage is entered. At 2:55, the system finally determines the new key control parameters: the damping coefficient is adjusted to 0.78, the motion angle is optimized to 35.2 degrees, and the natural frequency of the power generation system is adjusted to 0.92 Hz, which are immediately applied to the wave energy generation system.

[0231] At 2:57 after the application of the optimized parameters, the output power of the system quickly rises to 490kW, and the power fluctuation range is reduced to within ±10kW. At 3:00 in the morning, the energy conversion efficiency is improved to 47.5%, which is 6.2 percentage points higher than 41.3% at 2:30. At 3:05, the system analyzes the latest sea state data and finds that the average wind speed begins to gradually weaken to 5.4 meters / second, and the average wave period is extended to 7.8 seconds. The power generation system adjusts again according to the dynamic closed-loop feedback control mechanism, and finally stabilizes the power at around 500kW, with the energy loss cumulatively reduced to 9.5kWh in this period.

[0232] In order to further verify the adaptability of the application in extreme environmental changes, the test team conducted a sea state sudden change scene test at 10:00 on July 16, 2023. At 9:55 that morning, the sea state was stable, with an effective wave height of 2.2 meters, an average wave period of 7.9 seconds, and an average wind speed of 5.9 meters / second. The power output of the power generation system was stable at 510kW. However, at 10:02, monitoring point C (about 800 meters away from point B) detected that the average wind speed suddenly rose to 7.5 meters / second, the average current speed increased to 1.7 meters / second, and the average wave period shortened to 7.1 seconds. This mutation made the traditional control method unable to respond quickly in a short time, resulting in a sharp drop in system power to 450kW at 10:05, with energy loss reaching 8.3kWh in 3 minutes.

[0233] The method of the application started emergency optimization adjustment at 10:06, and the system automatically detected the environmental sudden change and immediately triggered the adaptive jellyfish search algorithm for local fine search. At 10:08, the optimization control system adjusted the key control parameters, optimized the motion angle of the wave energy conversion device to 37.5 degrees, adjusted the damping coefficient of the generator to 0.81, and completed the parameter update at 10:09, making the output power quickly rise to 495kW. At 10:12, the power stabilized at around 510kW, with a fluctuation range of ±8kW, which was significantly lower than the ±50kW fluctuation range of the traditional method. The energy conversion efficiency was improved to 48.2%.

[0234] In order to evaluate the long-term optimization effect of the present application, the test team conducted a 7-day comparative experiment on the traditional method and the method of the present application from July 10, 2023 to July 17, 2023, and the results are as follows:

[0235] The average power fluctuation range of the traditional method during the entire test period is ±45kW, and the energy conversion efficiency is stable at about 42.1%, and the total energy loss reaches 280kWh;

[0236] The average power fluctuation range of the method of the present application within 7 days is only ±12kW, the average energy conversion efficiency is 48.7%, and the total energy loss is controlled within 140kWh, which is reduced by 50% compared with the traditional method.

[0237] In addition, the test team selected 6 groups of sea state data for detailed comparison:

[0238] Table 1 Detailed comparison of 6 groups of sea state data between the present application and the traditional application

[0239]

[0240] From the above data, it can be seen that under the same sea state conditions, the power output of the method of the present application is increased by about 25kW on average compared with the traditional method, and the energy conversion efficiency is increased by about 6 percentage points. At the same time, the method of the present application has stronger adaptability to complex environmental changes, and can quickly optimize parameters when the sea state changes dramatically to ensure the stable operation of the power generation system and avoid power loss caused by sudden environmental changes.

[0241] The test data fully proves that the optimization control method of the present application can significantly improve the energy conversion efficiency of the wave energy power generation system in practical application, reduce energy loss, and provide more stable power output in the case of dramatic changes in marine environment, so that the wave energy power generation system has stronger environmental adaptability and higher operation stability, providing a more reliable and efficient technical solution for future large-scale ocean wave energy development.

[0242] The present application adopts a hybrid optimization strategy combining the bitter fish optimization algorithm and the adaptive jellyfish search algorithm. The traditional optimization method has the problems of low global search efficiency or local optimal disturbance, while the present application first uses the bitter fish optimization algorithm to perform global search on the entire parameter space, and quickly locates the candidate area of key parameters by adjusting the population distribution of the environment adaptively; Then the adaptive jellyfish search algorithm is used to perform local fine search on the candidate area, realizing accurate search and local refinement in the parameter space, ensuring the accuracy of the control parameters and the significant improvement of the system energy conversion efficiency.

[0243] The present application introduces environmental adaptability factors to optimize the initialization and dynamic adjustment of the population for real-time correction, so that the system can quickly respond and timely adjust the control parameters when facing sudden sea state changes, thereby maintaining high output power stability and low energy loss in complex and unstable marine environments, overcoming the defects of traditional methods that rely heavily on fixed parameters and lack of dynamic adaptability.

[0244] The present application designs a closed-loop feedback control mechanism based on real-time marine environmental data and the operating state of the power generation system. During the optimization process, the system continuously collects the latest environmental and equipment data, dynamically updates the key control parameters, forms a feedback loop, and realizes adaptive adjustment. The closed-loop mechanism can effectively deal with the variable interference factors in the marine environment, ensuring that the system is always in the best working state, thereby further improving the electric energy conversion efficiency and significantly reducing energy loss.

[0245] The above is only the preferred specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can make equivalent replacements or changes within the technical scope disclosed by the present application according to the technical solution and inventive concept of the present application, which should be covered within the protection scope of the present application.

Claims

1. A wave energy generation efficiency optimization control method based on artificial intelligence, characterized in that: The steps include: S1. Collect marine environment data sets; S2. Preprocessing the collected marine environment data set to form a preprocessed marine environment data set; S3. Constructing an optimization objective function based on the preprocessed marine environment data set and the structural parameters of the wave energy power generation system, and determining the corresponding constraints; S4. Using the bitterfish optimization algorithm to perform a global search for the optimization objective function, initialize the optimization population, and perform a rough search for key control parameters of the wave energy power generation system based on fitness evaluation to determine several candidate optimal parameter regions; S5. Use an adaptive jellyfish search algorithm to perform a localized, refined search of the candidate optimal parameter region, update and optimize key control parameters, and obtain key control parameters that further improve power conversion efficiency. S6. Apply key control parameters to the wave energy power generation system and, in combination with real-time monitored ocean environmental data and the operating status of the power generation system, form a dynamic closed-loop feedback control system to adjust the wave energy power generation system parameters in real time; The S4 includes: S41. Based on the optimization objective function and optimized variable sets Define the optimized population of environmentally adaptive bitterlings, set the parameter vectors of the optimized individuals, and adjust the population distribution according to the environmental dynamic characteristics of the wave energy generation system: ； in, Optimizing the population size for bitterlings, Indicates the The optimization variable vector for each individual: ； in, For the Active power output of each individual wave energy generation system, For the The conversion efficiency of individual wave energy power generation systems, For the The movement angle of each individual wave energy conversion device, For the The natural frequency of each individual wave energy device, For the The damping coefficient of each individual wave energy power generation device is calculated; the population is dynamically initialized and optimized based on the environmental characteristics of the wave energy power generation system, and the environmental impact factor is constructed based on the marine environmental data set. , through environmental factors Adjust the initial population distribution to adapt the initial population to the optimization calculation under different sea conditions; S42. Optimizing the population Each individual in ,Combined with the environmental adaptability of wave energy power generation system, an environmental constraint adaptive fitness function is constructed; S43. Based on the environment-adaptive bitterfish optimization strategy, a dynamic step size adjustment mechanism is introduced during the search process to update the optimization variables; S44. Adopt dynamic aggregation strategy to enhance the optimization ability of bitter fish group, introduce dynamic aggregation strategy to improve the convergence speed of bitter fish group, and define the attraction between bitter fish individuals ; S45. Iterate the S42-S44 process multiple times to finally determine the candidate optimal parameter region .

2. The method for optimizing and controlling wave energy generation efficiency based on artificial intelligence according to claim 1, characterized in that: Said S1 comprises the following steps: S11. Marine environment dataset defining wave energy generation systems , the marine environment data set includes significant wave height, average wave period, average wind speed, and average current speed; S12. Construct a time series set of environmental data based on the measurement requirements of the wave energy power generation system, and define the time series of the wave energy power generation system. Marine environment datasets ; S13. Use a multi-point measurement method to collect marine environmental datasets at different measurement points of the wave energy power generation system, and construct a marine environmental dataset with spatial distribution by combining the measurement points.

3. The method for optimizing and controlling wave energy generation efficiency based on artificial intelligence according to claim 1, characterized in that: The S2 comprises the following steps: S21. Marine environment dataset Perform integrity checks and eliminate incomplete data records to ensure that each data record contains effective wave height, average wave period, average wind speed, and average current speed, forming a complete marine environment data set; S22. Perform outlier detection on the marine environment dataset and remove data points that exceed the range to form a marine environment dataset after outlier removal; S23. Using a noise filtering method to smooth the marine environment dataset after outliers are removed, constructing a filtered marine environment dataset; S24. Perform normalization on the filtered marine environment dataset so that all environmental parameters are normalized to the same numerical range to form a normalized marine environment dataset. ; S25. Based on time step For the normalized marine environment dataset Resampling is performed to ensure that the time intervals of the data are consistent, forming the final preprocessed marine environment dataset.

4. The method for optimizing and controlling wave energy generation efficiency based on artificial intelligence according to claim 1, characterized in that: The S3 includes the following steps: S31. Based on the preprocessed marine environment dataset Extract characteristic variables related to the operation of wave energy power generation system and construct environmental parameter vector , based on the pre-processed marine environment data set and the structural parameters of the wave energy power generation system, establish the mapping relationship between environmental parameters and key performance variables of the system ; S32. Based on the optimization objectives of the wave energy power generation system, combined with the pre-processed marine environment dataset and optimized variable sets , establish the optimization objective function; S33. According to the operation constraints of the wave energy power generation system, set the optimization variable set Constraints; Set environmental adaptability constraints to ensure that the wave energy generation system remains stable in complex marine environments; S34. Based on the construction of the optimization objective function and the setting of the constraints, the final optimization solution is formed.

5. The method for optimizing and controlling wave energy generation efficiency based on artificial intelligence according to claim 1, characterized in that: The S5 comprises the following steps: S51. Based on the determined candidate optimal parameter region , construct the local search population of the adaptive jellyfish search algorithm and set the parameter vector of the local search individual; S52. For local search population Each individual in ,Adopt the adaptive jellyfish search algorithm to perform local fine search and update individual parameters; S53. According to the change of individual fitness during the local search process, the step factor Make adaptive adjustments; S54. Iterate steps S52 to S53 until the local search reaches the preset convergence condition or the maximum number of iterations is satisfied, and finally the key control parameters for further improving the power conversion efficiency are obtained.

6. An artificial intelligence-based wave energy generation efficiency optimization control device, characterized in that: The invention comprises a processor, a memory, and a wave energy power generation efficiency optimization control program stored in the memory and executable by the processor, wherein when the wave energy power generation efficiency optimization control program is executed by the processor, an artificial intelligence-based wave energy power generation efficiency optimization control method according to any one of claims 1 to 5 is implemented.

Images