Mountain wind energy resource dynamic evaluation method, device, equipment and medium

Abstract

The present application relates to the technical field of artificial intelligence, and provides a mountain wind energy resource dynamic evaluation method, device, equipment and medium, which can perform attention fusion on multi-dimensional initial data, so that the fused features take into account feature importance difference and comprehensiveness; a mountain wind energy evaluation data set with a wind energy power density label is constructed to provide a basis for supervised training of the model; the state space completely covers the core variables affecting the mountain wind energy evaluation, solving the problem of insufficient state space dimension of the traditional model and the inability to represent complex coupling relationships; the action space is non-static output, improving the self-adaptation ability of the model; the reward function solves the technical problem of being unable to simultaneously consider accuracy and efficiency; the double experience replay mechanism improves the evaluation accuracy of the model for complex wind field scenes; and the target network delay update strategy keeps the parameters of the target network stable within the iteration period, solving the problems of training shock and slow convergence caused by real-time fluctuations in target returns.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, equipment and medium for dynamic assessment of mountain wind energy resources. Background Technology

[0002] With the rapid development of the new energy industry, the development and utilization of wind energy, as a clean and renewable energy source, has become an important issue. Due to the complex terrain and irregular airflow in mountainous areas, the distribution of wind energy resources exhibits significant spatial heterogeneity and temporal dynamics. Accurate assessment of wind energy resources in mountainous areas is a core prerequisite for wind farm planning, turbine selection, and operation and maintenance optimization.

[0003] Currently, traditional methods for assessing wind energy resources in mountainous areas mainly include numerical simulation, field monitoring, and statistical analysis. Numerical simulation, based on fluid dynamics models, struggles to accurately depict the disturbances to airflow caused by complex mountain terrain (such as canyons and steep slopes), and its fixed model parameters cannot adapt to changes in dynamic factors such as seasons and weather. Field monitoring acquires actual data through the deployment of monitoring equipment, offering high accuracy, but it suffers from high monitoring costs, limited coverage, and poor data timeliness, making it impossible to achieve dynamic assessment across the entire region. Statistical analysis relies on historical data for fitting and prediction, neglecting dynamic factors such as subtle deformations in mountainous terrain and real-time weather fluctuations, resulting in significant assessment errors and failing to meet the needs of refined wind farm development.

[0004] In addition, existing assessment methods are mostly static assessments, which cannot adaptively adjust assessment strategies according to real-time environmental changes. This leads to discrepancies between the assessment results and the actual wind energy resources, thereby affecting the planning rationality and power generation efficiency of wind farms.

[0005] Therefore, there is an urgent need for a mountain wind energy resource assessment method that can take into account accuracy, timeliness and dynamic adaptability, in order to overcome the limitations of traditional methods. Summary of the Invention

[0006] In view of the above, it is necessary to provide a method, device, equipment and medium for dynamic assessment of mountain wind energy resources, in order to solve the problems of low accuracy, poor timeliness and lack of dynamic adaptability in the assessment of mountain wind energy resources.

[0007] A method for dynamic assessment of mountain wind energy resources, the method comprising:

[0008] In response to the instruction to dynamically assess the wind energy resources of the target mountain, the mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain are collected as multi-dimensional initial data.

[0009] Attention fusion is performed on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels;

[0010] A state space is constructed based on mountain terrain data and real-time environmental perception data that integrate historical wind energy monitoring data, an action space is constructed based on the evaluation strategy adjustment coefficients, a reward function is constructed based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and an initial model including a separate target network and a current network is constructed.

[0011] Based on the dual experience replay mechanism and the target network delayed update strategy, the initial model is trained using the mountain wind energy assessment dataset to obtain a dynamic assessment model for mountain wind energy resources.

[0012] According to the data dimensions of the multidimensional initial data, the multidimensional current data of the target mountain is collected in real time, and the multidimensional current data is input into the mountain wind energy resource dynamic assessment model to obtain the current wind energy power density.

[0013] The dynamic assessment results of wind energy resources in the target mountain area are generated based on the current wind power density.

[0014] A dynamic assessment device for mountain wind energy resources, the device comprising:

[0015] The data acquisition unit is used to respond to the command for dynamic assessment of wind energy resources of the target mountain by acquiring mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain as multi-dimensional initial data.

[0016] A fusion unit is used to perform attention fusion on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels.

[0017] The construction unit is used to construct a state space based on mountain terrain data and real-time environmental perception data that integrate historical wind energy characteristics from the historical wind energy monitoring data, construct an action space based on the evaluation strategy adjustment coefficient, construct a reward function based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and construct an initial model including a separate target network and a current network.

[0018] The training unit is used to train the initial model using the mountain wind energy assessment dataset based on the dual experience replay mechanism and the target network delayed update strategy, so as to obtain the mountain wind energy resource dynamic assessment model.

[0019] The input unit is used to collect multidimensional current data of the target mountain in real time according to the data dimensions of the multidimensional initial data, and input the multidimensional current data into the mountain wind energy resource dynamic assessment model to obtain the current wind energy power density.

[0020] The generation unit is used to generate a dynamic assessment result of the wind energy resources of the target mountain based on the current wind energy power density.

[0021] A computer device, the computer device comprising:

[0022] A memory for storing at least one instruction; and a processor for executing the instructions stored in the memory to implement the dynamic assessment method for mountain wind energy resources.

[0023] A computer-readable storage medium storing at least one instruction, which is executed by a processor in a computer device to implement the dynamic assessment method for mountain wind energy resources.

[0024] As can be seen from the above technical solutions, this invention can perform attention fusion on multi-dimensional initial data, making the fused features take into account both the differences in feature importance and comprehensiveness; it constructs a mountain wind energy assessment dataset with wind power density labels, providing a foundation for supervised training of the model; the state space fully covers the core variables affecting mountain wind energy assessment, solving the problem of insufficient state space dimension and inability to represent complex coupling relationships in traditional models; the action space is non-static, improving the model's adaptability; the reward function solves the technical problem of not being able to simultaneously balance accuracy and efficiency; the dual experience replay mechanism improves the model's assessment accuracy for complex wind field scenarios; the target network delayed update strategy keeps the target network parameters stable within the iteration cycle, solving the problems of training oscillation and slow convergence caused by real-time fluctuations in target rewards; and the mountain wind energy resource dynamic assessment model achieves efficient, accurate, and dynamically adaptable mountain wind energy resource dynamic assessment. Attached Figure Description

[0025] Figure 1 This is a flowchart of a preferred embodiment of the dynamic assessment method for mountain wind energy resources of the present invention;

[0026] Figure 2 This is a functional block diagram of a preferred embodiment of the dynamic assessment device for mountain wind energy resources of the present invention;

[0027] Figure 3 This is a schematic diagram of the computer device used in a preferred embodiment of the method for dynamic assessment of mountain wind energy resources according to the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] like Figure 1The diagram shown is a flowchart of a preferred embodiment of the dynamic assessment method for mountain wind energy resources of the present invention. The order of the steps in this flowchart can be changed, and some steps can be omitted, depending on different requirements.

[0030] The dynamic assessment method for mountain wind energy resources is applied to one or more computer devices. The computer device is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0031] The computer device can be any electronic product that can interact with the user, such as a personal computer, tablet computer, smartphone, personal digital assistant (PDA), game console, interactive network television (IPTV), smart wearable device, etc.

[0032] The computer equipment may also include network equipment and / or user equipment. The network equipment includes, but is not limited to, a single network server, a server group consisting of multiple network servers, or a cloud based on cloud computing consisting of a large number of hosts or network servers.

[0033] The server can be a standalone server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0034] Artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0035] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0036] The network in which the computer device is located includes, but is not limited to, the Internet, wide area network, metropolitan area network, local area network, and virtual private network (VPN).

[0037] S10, in response to the instruction to dynamically assess the wind energy resources of the target mountain, collect the mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain as multi-dimensional initial data.

[0038] In this embodiment, the dynamic assessment command for wind energy resources can be triggered by relevant personnel based on assessment needs.

[0039] In this embodiment, the mountain terrain data can be collected in the following manner:

[0040] First, the target mountainous area is divided into grid-like sub-regions, with the side length of each sub-region configurable to 100-200m. Coordinate points are marked at the boundary vertices and center of each sub-region. The coordinates can use a standard geodetic coordinate system, with a planar accuracy controlled within ±5cm. The overlap rate of coordinate points between adjacent sub-regions should be no less than 10% to ensure the continuity of the data stitching.

[0041] The coordinates of each sub-region are input into the UAV control system. The UAV, equipped with a 16-line or higher LiDAR (laser wavelength 905nm±5nm, point cloud density ≥50 points / ㎡) and a high-definition camera with at least 20 megapixels, scans the area according to the planned path. The flight path is dynamically planned based on the terrain complexity. Parallel flight paths are used in gentle areas (slope ≤15°) with a spacing of 20-30m; intersecting flight paths are used in steep areas (slope >15°) with a spacing reduced to 10-15m. The flight altitude is set according to a ground resolution of ≥2cm, and the relative altitude to the highest terrain point is no less than 50m to avoid blind spots in the LiDAR scan.

[0042] During the scanning process, point cloud data, GNSS (Global Navigation Satellite System) differential positioning data (positioning accuracy ≤3cm), and IMU (Inertial Measurement Unit) attitude data (attitude accuracy ≤0.1°) are recorded in real time. A data verification report is generated after each zone scan to ensure point cloud integrity ≥95% and no blurring or missing images. The acquired point cloud data undergoes denoising (e.g., removing noise points with a reflection intensity <200), coordinate transformation, and stitching (achieved using the ICP (Iterative Closest Point) algorithm with an error ≤3cm) to generate point cloud data in .las format. Terrain feature parameters such as mountain elevation, slope, aspect, and terrain roughness are then extracted.

[0043] In this embodiment, the historical wind energy monitoring data and the real-time environmental perception data can be collected in the following manner:

[0044] At least five wind energy monitoring stations will be deployed at different altitudes and slopes in the target mountain area. Each monitoring station will be equipped with wind speed sensors, wind direction sensors, air density sensors, temperature sensors, humidity sensors, etc., to collect historical wind energy data (including wind speed, wind direction, air density, turbulence intensity, etc.) for the past 3-10 years, with a sampling frequency of once every 10 minutes. Real-time environmental perception data will be collected in real time, with a sampling frequency of once every 5 minutes, including real-time wind speed, atmospheric temperature, humidity, air pressure, and data on minor topographic deformations (collected through displacement sensors).

[0045] Simultaneously, by linking regional meteorological and geological databases, the seasonal meteorological characteristics (such as winter wind speed distribution and the impact of summer precipitation on topography) and geological stability parameters (such as soil type and rock integrity) of the target mountain are supplemented to form the aforementioned multidimensional initial data.

[0046] The above embodiments can cover three core data sources: terrain, historical wind energy, and real-time environment. This solves the technical problem that a single data dimension cannot fully represent the distribution pattern of mountain wind energy resources, and provides full-dimensional and multi-scenario raw data support for subsequent model training, avoiding evaluation bias caused by data loss.

[0047] In this embodiment, in order to ensure data quality, the collected raw data also needs to be preprocessed.

[0048] Specifically, outlier removal, missing value interpolation, and standardization can be performed on the collected raw data. For example, the 3σ principle is used to remove outliers, that is, when the data value exceeds the mean ± 3 times the standard deviation, it is judged as an outlier and removed; missing values ​​are supplemented using linear interpolation based on time series to ensure data integrity; all data are standardized to map the data to the [0,1] interval to eliminate the influence of units.

[0049] S11, Attention fusion is performed on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels.

[0050] In this embodiment, the attention fusion of the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels includes:

[0051] Mountain terrain feature parameters, wind energy parameters, and environmental parameters are extracted from the multidimensional initial data.

[0052] An attention mechanism is used to perform correlation analysis based on the historical wind energy monitoring data to obtain a first weight corresponding to the mountain terrain feature parameters, a second weight corresponding to the wind energy parameters, and a third weight corresponding to the environmental parameters;

[0053] The first weight, the second weight, the third weight, the mountain terrain feature parameters, the wind energy parameters, and the environmental parameters are weighted and fused to obtain a fused feature vector;

[0054] The wind power density corresponding to the fused feature vector is determined based on the historical wind energy monitoring data, and the fused feature vector is labeled according to the wind power density corresponding to the fused feature vector to obtain the mountain wind energy assessment dataset.

[0055] The mountain terrain feature parameters may include elevation standard deviation (reflecting the degree of terrain undulation), mean slope (reflecting the steepness of the terrain), and aspect concentration (reflecting the uniformity of aspect distribution).

[0056] The wind energy parameters may include average wind speed, wind speed fluctuation rate (reflecting wind speed stability), and wind direction stability (reflecting the magnitude of wind direction change).

[0057] The environmental parameters may include temperature gradient, humidity change rate, and air pressure change.

[0058] When using the attention mechanism to perform correlation analysis based on the historical wind energy monitoring data, the attention mechanism can be used to calculate the Pearson correlation coefficient between each parameter and the wind energy power density label based on the historical wind energy monitoring data. After normalization, the first weight (e.g., 0.3) corresponding to the mountain terrain feature parameter, the second weight (e.g., 0.4) corresponding to the wind energy parameter, and the third weight (e.g., 0.3) corresponding to the environmental parameter are obtained.

[0059] Furthermore, the mountain wind energy assessment dataset can be split into a training set, a validation set, and a test set in a ratio of 7:2:1 for use in subsequent model training and validation.

[0060] In the above embodiments, the optimal weights of three types of features—terrain, wind energy, and environment—are automatically learned through an attention mechanism based on correlation analysis of historical wind energy monitoring data. This solves the problem that traditional equal-weighted fusion cannot reflect the differences in feature importance, significantly improving the effectiveness of feature representation. The fused feature vectors are labeled with historical measured wind power density, constructing a standardized dataset that meets the requirements of supervised learning. This addresses the technical pain point of lacking a unified labeled dataset for mountain wind energy assessment and provides accurate supervision signals for model training. Through weighted fusion, multi-dimensional original data is compressed into highly representative fused feature vectors, reducing data dimensionality while retaining core information, and also improving the efficiency and convergence speed of subsequent model training.

[0061] S12, construct a state space based on mountain terrain data and real-time environmental perception data that integrate historical wind energy characteristics from the historical wind energy monitoring data, construct an action space based on the evaluation strategy adjustment coefficient, construct a reward function based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and construct an initial model including a separate target network and a current network.

[0062] In this embodiment, the initial model can be a deep reinforcement learning model. Constructing an initial model that separates the target network from the current network solves the problem of frequent fluctuations in target values ​​and training oscillations in traditional DQN (Deep Q-Network) training from an architectural perspective, providing architectural guarantees for subsequent stable training.

[0063] The initial model's network structure may include an input layer, a feature extraction layer, a fully connected layer, and an output layer. The input layer receives the fused feature vector, with a dimension of n×1 (n is the total number of feature parameters; in this embodiment, n=9, including 3 terrain features, 3 wind energy features, and 3 environmental features). The feature extraction layer uses a convolutional neural network to extract deep correlation information from the features, configuring two convolutional layers with kernel sizes of 3×1 and 2×1, and using ReLU activation. The fully connected layer has two hidden layers with 64 and 32 neurons, respectively, and uses ReLU activation. The output layer outputs the Q-value (reward) of the action space, corresponding to the gain of adjusting coefficients for different evaluation strategies.

[0064] The state space includes a concatenated vector of mountain terrain feature vector and real-time environmental parameter vector, such as state space S = [mountain terrain feature vector, environmental parameter vector], with a dimension of 9×1.

[0065] The state space integrates historical wind energy characteristics, terrain and environmental parameters, fully covering the core variables affecting mountain wind energy assessment, and solving the problems of insufficient state space dimension and inability to represent complex coupling relationships in traditional models.

[0066] The action space is a set of wind speed correction coefficient, terrain influence coefficient, and environmental compensation coefficient; the value range of each coefficient can be [0.8, 1.2], and a total of 125 discrete actions are divided (each coefficient takes 5 discrete values).

[0067] The action space enables the model to dynamically optimize the evaluation results, rather than outputting statically, thus improving the model's adaptability.

[0068] The formula for the reward function is: R = α × (1 - |E - E0| / E0) + β × (1 - T / T0);

[0069] Where R represents the reward function value; α represents the evaluation accuracy weight coefficient; β represents the data timeliness weight coefficient; α+β=1, such as α=0.7, β=0.3; E represents the dynamic evaluation wind power density output by the model; E0 represents the actual monitored wind power density; T represents the model evaluation time; T0 represents the preset maximum allowable time, such as T0=1s.

[0070] The larger the reward function value, the better the evaluation strategy.

[0071] The reward function takes the accuracy error and timeliness of historical data as its core, and integrates the dual objectives of accuracy and real-time performance into model training, thus solving the technical problem that traditional models cannot simultaneously achieve both accuracy and efficiency.

[0072] In the above embodiments, the state space and reward function are both based on historical wind energy monitoring data as the core calibration basis, so that the model fits the actual wind field pattern of the target mountain from the construction stage, avoiding the problem of poor scene adaptability of general models.

[0073] S13, Based on the dual experience replay mechanism and the target network delayed update strategy, the initial model is trained using the mountain wind energy assessment dataset to obtain the mountain wind energy resource dynamic assessment model.

[0074] In this embodiment, the step of training the initial model using the mountain wind energy assessment dataset based on the dual experience replay mechanism and the target network delayed update strategy to obtain the dynamic assessment model of mountain wind energy resources includes:

[0075] During training, each generated evaluation sample, including state, action, reward, and next state, is stored in a low-priority buffer.

[0076] The evaluation error (the relative error between the model evaluation value E and the actual monitoring value E0) of each evaluation sample in the low-priority buffer is calculated in real time, and evaluation samples with evaluation errors greater than the error threshold are moved from the low-priority buffer to the high-priority buffer.

[0077] During model training sampling, samples are preferentially drawn from the high-priority buffer to train the initial model;

[0078] When the iteration termination condition is met, training is stopped and the target network is discarded, while the current network with optimal parameters is retained as the dynamic evaluation model for mountain wind energy resources.

[0079] The sampling proportion of the high-priority buffer is greater than that of the low-priority buffer;

[0080] The target network is used to provide a stable target reward for the current network during training; after each preset number of iterations, the latest parameters of the current network are fully synchronized to the target network.

[0081] The evaluation error is the relative error between the model's evaluated value and the actual monitored value. For example, evaluation samples with a relative error greater than 5% can be moved from the low-priority buffer to the high-priority buffer, where the high-priority buffer only stores high-error samples.

[0082] In particular, during model training sampling, samples are preferentially drawn from the high-priority buffer to train the initial model, which can improve the model's learning efficiency for high-error samples.

[0083] The low-priority buffer and the high-priority buffer can follow the first-in, first-out principle. For example, when the total storage capacity of the buffer reaches a certain threshold, the earliest stored sample is removed to ensure the timeliness of the buffer data and adapt to the dynamic changes of the mountain wind field.

[0084] During sampling training, samples from the two buffers are mixed and input, and the sampling ratio of high-error samples is not less than a certain proportion, such as 60%, so as to ensure that the model focuses on learning and evaluating scenarios with large deviations.

[0085] For example, when training the model, the training set of the aforementioned mountain wind energy assessment dataset can be input into the initial model, setting the initial learning rate to 0.001, the number of iterations to 10,000, the experience replay buffer size to 100,000, and the target network update frequency to once every 100 iterations. This avoids excessive fluctuations in the target value during training and improves the model's convergence speed and stability. During training, an ε-greedy strategy is used to select actions, with the ε value linearly decreasing from 0.9 to 0.1. That is, in the early stage of training, ε=0.9, exploring new actions with a 90% probability and selecting the current optimal action with a 10% probability; in the later stage of training, ε=0.1, exploring new actions with a 10% probability and selecting the current optimal action with a 90% probability, to balance the model's exploration and utilization.

[0086] Furthermore, the trained model is validated using a validation set, and metrics such as evaluation accuracy (the goodness of fit between the evaluated value and the actual value) and evaluation time are calculated. In this embodiment, the preset evaluation accuracy threshold is ≥90%, and the evaluation time threshold is ≤1s. If the model evaluation accuracy is lower than 90% or the time exceeds 1s, the model network structure (such as increasing the number of neurons in the convolutional layer) and hyperparameters (such as adjusting the learning rate and ε decay rate) are adjusted, and retraining is performed.

[0087] Furthermore, the qualified model is tested using a test set. The evaluation accuracy, time consumption, and error distribution of the test set are calculated. If the evaluation accuracy of the test set is ≥90%, the time consumption is ≤1s, and the error distribution is uniform (the proportion of samples with relative error ≤5% is ≥95%), then the mountain wind energy evaluation dataset is determined. Otherwise, the model parameters are adjusted, and the training and validation steps are repeated until the performance requirements are met.

[0088] The dual experience replay mechanism enables priority sampling and training of high-error samples, solving the problems of insufficient learning of difficult samples and poor model generalization in traditional experience replay, and significantly improving the model's evaluation accuracy for complex wind field scenarios.

[0089] By employing the target network delayed update strategy, a fixed-frequency delayed update of the target network is adopted, ensuring parameter stability within the iteration cycle. This solves the problems of training oscillation and slow convergence caused by real-time fluctuations in the target value in traditional DQN, significantly improving the stability and convergence speed of model training. Periodic synchronization of dual network parameters ensures the reference value of the target network while allowing the current network to continuously learn the latest patterns, achieving a balance between stable training and continuous optimization.

[0090] By combining the dual experience replay mechanism and the target network delayed update strategy, the synergistic optimization of the two mechanisms is achieved, enabling the trained model to have high evaluation accuracy, strong generalization and high stability. It can adapt to the complex wind field scenario of the target mountain and provide core model support for subsequent real-time dynamic evaluation.

[0091] In this embodiment, after obtaining the dynamic assessment model of mountain wind energy resources, the method further includes:

[0092] The actual wind power density monitoring data of the target mountain area is collected at preset time intervals;

[0093] Obtain the wind power density assessment data output by the dynamic assessment model of mountain wind energy resources;

[0094] Calculate the relative error between the wind power density assessment data and the actual wind power density monitoring data;

[0095] When the relative error is greater than a preset threshold, the current assessment accuracy and current assessment time of the mountain wind energy resource dynamic assessment model are obtained.

[0096] When the current evaluation accuracy is greater than the accuracy threshold, the evaluation accuracy weight coefficient is increased by a first preset step size; or when the current evaluation time is greater than the time consumption threshold, the data timeliness weight coefficient is increased by a second preset step size.

[0097] The mountain wind energy assessment dataset will be supplemented based on the latest monitoring data.

[0098] Based on the increased assessment accuracy weight coefficient or the increased data timeliness weight coefficient, the mountain wind energy assessment dataset is used to incrementally train the dynamic assessment model of mountain wind energy resources.

[0099] For example, actual wind energy monitoring data can be collected every 5 minutes, and the relative error between the dynamic assessment result and the actual monitoring data can be calculated. When the relative error exceeds 5%, a feedback optimization mechanism is triggered. Furthermore, the reward function weights α and β of the model are adjusted based on the error value. If the assessment accuracy error is large, the value of α is increased (e.g., adjusted to 0.8) to enhance the model's emphasis on assessment accuracy; if the assessment takes too long, the value of β is increased (e.g., adjusted to 0.4) to improve the model's timeliness. Simultaneously, the latest real-time monitoring data and terrain dynamic data are added to the mountain wind energy assessment dataset, and the model is incrementally trained without retraining the entire model, effectively reducing training time.

[0100] In this embodiment, after obtaining the dynamic assessment model of mountain wind energy resources, the method further includes:

[0101] The performance of the dynamic evaluation model for mountain wind energy resources is tested at preset intervals, and the model is continuously optimized based on the test results.

[0102] For example, the model's performance can be tested every 24 hours to calculate the average accuracy and average time of the evaluation results throughout the day. If the average accuracy is below 90% or the average time exceeds 1 second, the model's network structure and hyperparameters can be further adjusted to continuously optimize the model's performance, ensuring that the model can adapt to the influence of dynamic factors such as small deformations of mountainous terrain and seasonal weather changes, and maintain high evaluation accuracy and timeliness in the long term.

[0103] S14. Collect multidimensional current data of the target mountain in real time according to the data dimensions of the multidimensional initial data, and input the multidimensional current data into the mountain wind energy resource dynamic assessment model to obtain the current wind energy power density.

[0104] In this embodiment, the mountain terrain data (such as minor terrain deformation and vegetation cover changes) and real-time environmental perception data (such as real-time wind speed, temperature, humidity, etc.) of the target mountain can be collected in real time. The real-time fused feature vector is generated according to the preprocessing and feature fusion method described above and input into the mountain wind energy resource dynamic assessment model.

[0105] The model adjusts the coefficients by selecting the optimal evaluation strategy based on the real-time feature vectors through reinforcement learning, and outputs the current wind power density.

[0106] S15, Generate a dynamic assessment result of wind energy resources for the target mountain based on the current wind energy power density.

[0107] In this embodiment, generating the dynamic assessment result of wind energy resources of the target mountain based on the current wind power density includes:

[0108] The current wind power density is classified according to the wind energy resource classification strategy to obtain the classification results.

[0109] A dynamic assessment heat map of mountain wind energy resources is generated based on the classification results.

[0110] The visualized mountain wind energy resource dynamic assessment heat map includes wind energy resource levels for different regions.

[0111] For example, the wind energy resource level can be divided into four levels: excellent, good, medium, and poor, corresponding to wind power densities of ≥600W / m³. 2 400-600W / m 2 200-400W / m 2 <200W / m 2 Furthermore, the wind energy resource levels are visualized to generate a dynamic assessment heat map of mountain wind energy resources, marking the wind energy resource potential levels of different regions. This visually presents the spatial distribution and dynamic changes of mountain wind energy resources, providing a visual reference for wind farm planning.

[0112] In this embodiment, the available wind energy duration, wind energy stability coefficient, and extreme wind speed probability for different areas of the target mountain can also be derived based on the current wind power density directly output by the model, combined with preset industry algorithm rules and statistical rules, including:

[0113] (1) Utilizable wind energy duration: Based on real-time wind speed (fusion characteristics), power density threshold (industry standard: the lower limit of power density that can achieve effective power generation), and combined with the time distribution pattern of historical wind speed (which can be obtained based on historical wind energy monitoring data), it is statistically derived;

[0114] (2) Wind energy stability coefficient: Based on the two features of wind speed fluctuation rate and wind direction stability learned by the model, it is calculated according to the preset weight (consistent with the feature weight during training). The value range can be mapped to [0,1]. The closer the coefficient is to 1, the more stable the wind energy is.

[0115] (3) Extreme wind speed probability: Based on the frequency of extreme wind speed (≥25m / s) in historical wind energy monitoring data, combined with the changing trend of real-time environmental parameters (such as air pressure and temperature gradient), it is derived through the probability statistics module of the model and used as a reference for wind farm unit selection.

[0116] This embodiment deeply integrates deep reinforcement learning with the assessment of mountain wind energy resources, breaking through the limitations of traditional static assessment and constructing an assessment model that can adapt to dynamic changes in mountain terrain and environment. Furthermore, by accurately acquiring various types of data, optimizing the model structure, and introducing dynamic reward functions and feedback optimization mechanisms, it achieves accurate, real-time, and dynamic assessment of mountain wind energy resources. This solves the technical problems of low accuracy, poor timeliness, and inability to adapt to dynamic changes inherent in traditional methods, providing a scientific and reliable basis for the planning, turbine selection, and operation and maintenance optimization of mountain wind farms, and has significant engineering application value.

[0117] As can be seen from the above technical solutions, this invention can perform attention fusion on multi-dimensional initial data, making the fused features take into account both the differences in feature importance and comprehensiveness; it constructs a mountain wind energy assessment dataset with wind power density labels, providing a foundation for supervised training of the model; the state space fully covers the core variables affecting mountain wind energy assessment, solving the problem of insufficient state space dimension and inability to represent complex coupling relationships in traditional models; the action space is non-static, improving the model's adaptability; the reward function solves the technical problem of not being able to simultaneously balance accuracy and efficiency; the dual experience replay mechanism improves the model's assessment accuracy for complex wind field scenarios; the target network delayed update strategy keeps the target network parameters stable within the iteration cycle, solving the problems of training oscillation and slow convergence caused by real-time fluctuations in target rewards; and the mountain wind energy resource dynamic assessment model achieves efficient, accurate, and dynamically adaptable mountain wind energy resource dynamic assessment.

[0118] like Figure 2 The diagram shown is a functional block diagram of a preferred embodiment of the dynamic assessment device for mountain wind energy resources of the present invention. The dynamic assessment device 11 for mountain wind energy resources includes a data acquisition unit 110, a fusion unit 111, a construction unit 112, a training unit 113, an input unit 114, and a generation unit 115. The module / unit referred to in this invention is a series of computer program segments that can be executed by a processor and perform a fixed function, and are stored in memory. In this embodiment, the functions of each module / unit will be described in detail in subsequent embodiments.

[0119] The acquisition unit 110 is used to collect mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain as multi-dimensional initial data in response to the dynamic assessment command of wind energy resources of the target mountain.

[0120] The fusion unit 111 is used to perform attention fusion on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels.

[0121] The construction unit 112 is used to construct a state space based on mountain terrain data and real-time environmental perception data that integrate historical wind energy characteristics from the historical wind energy monitoring data, construct an action space based on the evaluation strategy adjustment coefficient, construct a reward function based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and construct an initial model including a separate target network and a current network.

[0122] The training unit 113 is used to train the initial model using the mountain wind energy assessment dataset based on the dual experience replay mechanism and the target network delayed update strategy, so as to obtain a dynamic assessment model of mountain wind energy resources.

[0123] The input unit 114 is used to collect multidimensional current data of the target mountain in real time according to the data dimensions of the multidimensional initial data, and input the multidimensional current data into the mountain wind energy resource dynamic evaluation model to obtain the current wind energy power density.

[0124] The generation unit 115 is used to generate a dynamic assessment result of the wind energy resources of the target mountain based on the current wind energy power density.

[0125] As can be seen from the above technical solutions, this invention can perform attention fusion on multi-dimensional initial data, making the fused features take into account both the differences in feature importance and comprehensiveness; it constructs a mountain wind energy assessment dataset with wind power density labels, providing a foundation for supervised training of the model; the state space fully covers the core variables affecting mountain wind energy assessment, solving the problem of insufficient state space dimension and inability to represent complex coupling relationships in traditional models; the action space is non-static, improving the model's adaptability; the reward function solves the technical problem of not being able to simultaneously balance accuracy and efficiency; the dual experience replay mechanism improves the model's assessment accuracy for complex wind field scenarios; the target network delayed update strategy keeps the target network parameters stable within the iteration cycle, solving the problems of training oscillation and slow convergence caused by real-time fluctuations in target rewards; and the mountain wind energy resource dynamic assessment model achieves efficient, accurate, and dynamically adaptable mountain wind energy resource dynamic assessment.

[0126] like Figure 3 The diagram shown is a schematic representation of the computer equipment used in a preferred embodiment of the method for dynamic assessment of mountain wind energy resources according to the present invention.

[0127] The computer device 1 may include a memory 12, a processor 13, and a bus (the arrow in the figure represents the bus), and may also include a computer program stored in the memory 12 and capable of running on the processor 13, such as a dynamic assessment program for mountain wind energy resources.

[0128] Those skilled in the art will understand that the schematic diagram is merely an example of computer device 1 and does not constitute a limitation on computer device 1. Computer device 1 can be either a bus topology or a star topology. Computer device 1 may also include more or fewer other hardware or software than shown in the diagram, or different component arrangements. For example, computer device 1 may also include input / output devices, network access devices, etc.

[0129] It should be noted that the computer device 1 described is merely an example. Other existing or future electronic products that are adaptable to this invention should also be included within the scope of protection of this invention and are incorporated herein by reference.

[0130] The memory 12 includes at least one type of readable storage medium, such as flash memory, portable hard drive, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 12 can be an internal storage unit of the computer device 1, such as a portable hard drive of the computer device 1. In other embodiments, the memory 12 can be an external storage device of the computer device 1, such as a plug-in portable hard drive, smart media card (SMC), secure digital card (SD), flash card, etc., equipped on the computer device 1. Furthermore, the memory 12 can include both internal and external storage units of the computer device 1. The memory 12 can be used not only to store application software and various types of data installed on the computer device 1, such as the code of a dynamic assessment program for mountain wind energy resources, but also to temporarily store data that has been output or will be output.

[0131] In some embodiments, the processor 13 may be composed of integrated circuits, such as a single packaged integrated circuit or multiple integrated circuits with the same or different functions, including combinations of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips. The processor 13 is the control unit of the computer device 1, connecting various components of the computer device 1 via various interfaces and lines. It executes programs or modules stored in the memory 12 (e.g., executing a dynamic assessment program for mountain wind energy resources) and calls data stored in the memory 12 to perform various functions of the computer device 1 and process data.

[0132] The processor 13 executes the operating system of the computer device 1 and various installed applications. The processor 13 executes these applications to implement the steps in the various embodiments of the dynamic assessment method for mountain wind energy resources described above, for example... Figure 1 The steps are shown.

[0133] For example, the computer program may be divided into one or more modules / units, which are stored in the memory 12 and executed by the processor 13 to complete the present invention. The one or more modules / units may be a series of computer-readable instruction segments capable of performing specific functions, which describe the execution process of the computer program in the computer device 1. For example, the computer program may be divided into an acquisition unit 110, a fusion unit 111, a construction unit 112, a training unit 113, an input unit 114, and a generation unit 115.

[0134] The integrated unit implemented as a software functional module described above can be stored in a computer-readable storage medium. This software functional module, stored in a storage medium, includes several instructions to cause a computer device (which may be a personal computer, computer equipment, or network device, etc.) or processor to execute portions of the dynamic assessment method for mountain wind energy resources described in the various embodiments of this invention.

[0135] If the modules / units integrated in the computer device 1 are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware devices. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above.

[0136] The computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory, etc.

[0137] Furthermore, the computer-readable storage medium may primarily include a stored program area and a stored data area, wherein the stored program area may store the operating system, an application program required for at least one function, etc.; and the stored data area may store data created based on the use of blockchain nodes, etc.

[0138] The blockchain referred to in this invention is a novel application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked together using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include an underlying blockchain platform, a platform product service layer, and an application service layer.

[0139] The bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, in... Figure 3 The bus is represented by only one straight line, but this does not mean that there is only one bus or one type of bus. The bus is configured to enable communication between the memory 12 and at least one processor 13, etc.

[0140] Although not shown, the computer device 1 may also include a power supply (such as a battery) to power various components. Preferably, the power supply can be logically connected to the at least one processor 13 through a power management device, thereby enabling functions such as charging management, discharging management, and power consumption management. The power supply may also include one or more DC or AC power supplies, recharging devices, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components. The computer device 1 may also include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be described in detail here.

[0141] Furthermore, the computer device 1 may also include a network interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a Wi-Fi interface, a Bluetooth interface, etc.), which is typically used to establish communication connections between the computer device 1 and other computer devices.

[0142] Optionally, the computer device 1 may further include a user interface, which may be a display, an input unit (such as a keyboard), and optionally, a standard wired interface or a wireless interface. Optionally, in some embodiments, the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen, etc. The display may also be appropriately referred to as a screen or display unit, used to display information processed in the computer device 1 and to display a visual user interface.

[0143] It should be understood that the embodiments described are for illustrative purposes only and are not limited to this structure in the scope of the patent application.

[0144] It will be understood by those skilled in the art that Figure 3 The structure shown does not constitute a limitation on the computer device 1, and may include fewer or more components than shown, or combine certain components, or have different component arrangements.

[0145] Combination Figure 1 The memory 12 in the computer device 1 stores multiple instructions to implement a dynamic assessment method for mountain wind energy resources, and the processor 13 can execute the multiple instructions to achieve the following:

[0146] In response to the instruction to dynamically assess the wind energy resources of the target mountain, the mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain are collected as multi-dimensional initial data.

[0147] Attention fusion is performed on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels;

[0148] A state space is constructed based on mountain terrain data and real-time environmental perception data that integrate historical wind energy monitoring data, an action space is constructed based on the evaluation strategy adjustment coefficients, a reward function is constructed based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and an initial model including a separate target network and a current network is constructed.

[0149] Based on the dual experience replay mechanism and the target network delayed update strategy, the initial model is trained using the mountain wind energy assessment dataset to obtain a dynamic assessment model for mountain wind energy resources.

[0150] According to the data dimensions of the multidimensional initial data, the multidimensional current data of the target mountain is collected in real time, and the multidimensional current data is input into the mountain wind energy resource dynamic assessment model to obtain the current wind energy power density.

[0151] The dynamic assessment results of wind energy resources in the target mountain area are generated based on the current wind power density.

[0152] Specifically, the processor 13's implementation method for the above instructions can be found in [reference needed]. Figure 1 The descriptions of the relevant steps in the corresponding embodiments are not repeated here.

[0153] It should be noted that all the data involved in this case was legally obtained.

[0154] If any AI models, software tools, or components not belonging to this company appear in the embodiments of this invention, they are merely illustrative examples and do not represent actual use. All user personal information involved in the embodiments of this invention has been obtained by an entity authorized (with the knowledge and consent) or fully authorized by all parties through various legal and compliant means. The collection, storage, use, processing, transmission, provision, and disclosure of the information, data, and signals involved all comply with relevant laws and regulations and do not violate public order and good morals.

[0155] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0156] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0157] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0158] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0159] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0160] Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be embraced within the invention. No appended diagram markings in the claims should be construed as limiting the scope of the claims.

[0161] Furthermore, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices described in this invention can also be implemented by a single unit or device through software or hardware. Terms such as "first," "second," etc., are used to indicate names and do not indicate any specific order.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for dynamic assessment of mountain wind energy resources, characterized in that, The dynamic assessment method for mountain wind energy resources includes: In response to the instruction to dynamically assess the wind energy resources of the target mountain, the mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain are collected as multi-dimensional initial data. Attention fusion is performed on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels; A state space is constructed based on mountain terrain data and real-time environmental perception data that integrate historical wind energy monitoring data, an action space is constructed based on the evaluation strategy adjustment coefficients, a reward function is constructed based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and an initial model including a separate target network and a current network is constructed. Based on the dual experience replay mechanism and the target network delayed update strategy, the initial model is trained using the mountain wind energy assessment dataset to obtain a dynamic assessment model for mountain wind energy resources. According to the data dimensions of the multidimensional initial data, the multidimensional current data of the target mountain is collected in real time, and the multidimensional current data is input into the mountain wind energy resource dynamic assessment model to obtain the current wind energy power density. The dynamic assessment results of wind energy resources in the target mountain area are generated based on the current wind power density.

2. The method for dynamic assessment of mountain wind energy resources as described in claim 1, characterized in that, The attention fusion of the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels includes: Mountain terrain feature parameters, wind energy parameters, and environmental parameters are extracted from the multidimensional initial data. An attention mechanism is used to perform correlation analysis based on the historical wind energy monitoring data to obtain a first weight corresponding to the mountain terrain feature parameters, a second weight corresponding to the wind energy parameters, and a third weight corresponding to the environmental parameters; The first weight, the second weight, the third weight, the mountain terrain feature parameters, the wind energy parameters, and the environmental parameters are weighted and fused to obtain a fused feature vector; The wind power density corresponding to the fused feature vector is determined based on the historical wind energy monitoring data, and the fused feature vector is labeled according to the wind power density corresponding to the fused feature vector to obtain the mountain wind energy assessment dataset.

3. The method for dynamic assessment of mountain wind energy resources as described in claim 1, characterized in that: The state space includes a concatenated vector of mountain terrain feature vectors and real-time environmental parameter vectors; The action space is a set of wind speed correction coefficient, terrain influence coefficient, and environmental compensation coefficient; The formula for the reward function is: R = α × (1 - |E - E0| / E0) + β × (1 - T / T0); Where R represents the reward function value; α represents the evaluation accuracy weight coefficient; β represents the data timeliness weight coefficient; α+β=1; E represents the dynamic evaluation wind power density output by the model; E0 represents the actual monitored wind power density; T represents the model evaluation time; T0 represents the preset maximum allowable time.

4. The method for dynamic assessment of mountain wind energy resources as described in claim 1, characterized in that, The method based on the dual experience replay mechanism and the target network delayed update strategy, using the mountain wind energy assessment dataset to train the initial model, yields the following dynamic assessment model for mountain wind energy resources: During training, each generated evaluation sample, including state, action, reward, and next state, is stored in a low-priority buffer. The evaluation error of each evaluation sample in the low-priority buffer is calculated in real time, and evaluation samples with evaluation errors greater than the error threshold are moved from the low-priority buffer to the high-priority buffer. During model training sampling, samples are preferentially drawn from the high-priority buffer to train the initial model; When the iteration termination condition is met, training is stopped and the target network is discarded, while the current network with optimal parameters is retained as the dynamic evaluation model for mountain wind energy resources. The sampling proportion of the high-priority buffer is greater than that of the low-priority buffer; The target network is used to provide a stable target reward for the current network during training; after each preset number of iterations, the latest parameters of the current network are fully synchronized to the target network.

5. The method for dynamic assessment of mountain wind energy resources as described in claim 3, characterized in that, After obtaining the dynamic assessment model for mountain wind energy resources, the method further includes: The actual wind power density monitoring data of the target mountain area is collected at preset time intervals; Obtain the wind power density assessment data output by the dynamic assessment model of mountain wind energy resources; Calculate the relative error between the wind power density assessment data and the actual wind power density monitoring data; When the relative error is greater than a preset threshold, the current assessment accuracy and current assessment time of the mountain wind energy resource dynamic assessment model are obtained. When the current evaluation accuracy is greater than the accuracy threshold, the evaluation accuracy weight coefficient is increased by a first preset step size; or when the current evaluation time is greater than the time consumption threshold, the data timeliness weight coefficient is increased by a second preset step size. The mountain wind energy assessment dataset will be supplemented based on the latest monitoring data. Based on the increased assessment accuracy weight coefficient or the increased data timeliness weight coefficient, the mountain wind energy assessment dataset is used to incrementally train the dynamic assessment model of mountain wind energy resources.

6. The method for dynamic assessment of mountain wind energy resources as described in claim 1, characterized in that, After obtaining the dynamic assessment model for mountain wind energy resources, the method further includes: The performance of the dynamic evaluation model for mountain wind energy resources is tested at preset intervals, and the model is continuously optimized based on the test results.

7. The method for dynamic assessment of mountain wind energy resources as described in claim 1, characterized in that, The process of generating dynamic assessment results of wind energy resources for the target mountain based on the current wind power density includes: The current wind power density is classified according to the wind energy resource classification strategy to obtain the classification results. A dynamic assessment heat map of mountain wind energy resources is generated based on the classification results. The visualized mountain wind energy resource dynamic assessment heat map includes wind energy resource levels for different regions.

8. A dynamic assessment device for mountain wind energy resources, characterized in that, The mountain wind energy resource dynamic assessment device includes: The data acquisition unit is used to respond to the command for dynamic assessment of wind energy resources of the target mountain by acquiring mountain terrain data, historical wind energy monitoring data and real-time environmental perception data of the target mountain as multi-dimensional initial data. A fusion unit is used to perform attention fusion on the multidimensional initial data to construct a mountain wind energy assessment dataset with wind power density labels. The construction unit is used to construct a state space based on mountain terrain data and real-time environmental perception data that integrate historical wind energy characteristics from the historical wind energy monitoring data, construct an action space based on the evaluation strategy adjustment coefficient, construct a reward function based on the evaluation accuracy error and data timeliness calibrated based on the historical wind energy monitoring data, and construct an initial model including a separate target network and a current network. The training unit is used to train the initial model using the mountain wind energy assessment dataset based on the dual experience replay mechanism and the target network delayed update strategy, so as to obtain the mountain wind energy resource dynamic assessment model. The input unit is used to collect multidimensional current data of the target mountain in real time according to the data dimensions of the multidimensional initial data, and input the multidimensional current data into the mountain wind energy resource dynamic assessment model to obtain the current wind energy power density. The generation unit is used to generate a dynamic assessment result of the wind energy resources of the target mountain based on the current wind energy power density.

9. A computer device, characterized in that, The computer device includes: A memory for storing at least one instruction; and a processor for executing the instructions stored in the memory to implement the dynamic assessment method for mountain wind energy resources as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores at least one instruction, which is executed by a processor in a computer device to implement the dynamic assessment method for mountain wind energy resources as described in any one of claims 1 to 7.

Images