Simulation system and simulation method for real-time emergency disposal of wind power plant

Abstract

The invention belongs to the technical field of wind power generation, and discloses a real-time emergency disposal simulation system and simulation method for a wind power plant, and the system comprises an edge calculation module which is disposed on each fan of the wind power plant and is used for executing real-time data preprocessing and preliminary fault detection; the central processing module is used for executing deep analysis and global decision; the virtual reality module is used for generating a training scene and providing virtual training; the augmented reality module is used for providing real-time guidance in an actual environment; the quantum computing module is used for solving the resource scheduling optimization problem; and the biosensing module is used for collecting physiological data of an operator. Training of various real and complex abnormal conditions can be achieved, adjustment can be conducted according to personnel feedback, and the problems of rapid fault diagnosis of the wind power plant in the complex environment, emergency resource optimization scheduling under the extreme weather condition and cognitive load management of operators in the high-voltage environment are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of wind power generation, and in particular relates to a real-time emergency handling simulation system and a simulation method for a wind farm. Background Art

[0002] Wind farm emergency response refers to a series of activities that take timely and effective measures to ensure personnel safety, equipment integrity and stable operation of the power system in response to sudden failures, accidents or abnormal conditions during the operation of wind power equipment. These activities include fault detection, diagnosis, resource scheduling, on-site response and personnel training. Effective emergency response is of great significance to reducing wind farm operation risks, improving power generation efficiency and extending equipment service life.

[0003] At present, emergency response in wind farms mainly relies on manual experience and preset emergency plans. Traditional fault detection methods are often slow to respond and it is difficult to identify and locate complex faults in a timely manner, especially under severe weather conditions. Resource scheduling usually adopts static optimization algorithms, which are difficult to adapt to the rapidly changing wind farm environment. In addition, the training of operators mainly relies on theoretical learning and limited field exercises, which makes it difficult to fully simulate various complex emergency scenarios. These limitations lead to problems such as slow emergency response, low decision-making accuracy, and poor training results. Therefore, there is an urgent need for a real-time emergency response system for wind farms that can achieve rapid fault diagnosis, intelligent resource scheduling, and efficient personnel training. The system should be able to quickly and accurately detect and diagnose various faults in a complex wind farm environment, optimize the allocation of emergency resources, and provide operators with a realistic simulation training environment, thereby significantly improving the overall emergency response capabilities of wind farms.

[0004] The Chinese patent publication number is CN117875941A, and the name is a patent application for a wind power operation and maintenance mixed reality interactive training system. The system includes a data center for collecting, storing, processing and analyzing wind farm data, providing data support and intelligent services for the training center and the intelligent center; a mixed reality center for providing mixed reality scenes and mixed reality content for the training center; a training center for organizing, managing and evaluating the training activities of wind power operation and maintenance personnel; and an intelligent center for intelligently analyzing and recommending wind farm data and wind power operation and maintenance personnel's behaviors, providing intelligent solutions and personalized learning paths for wind power operation and maintenance based on the real-time status of the wind farm and the personalized needs of the trainees, and improving the efficiency and quality of wind power operation and maintenance. This patent application can perform simulation training, but cannot achieve training for various real and complex abnormal situations and cannot be adjusted based on personnel feedback. Summary of the invention

[0005] In order to overcome the problems existing in the above-mentioned prior art, the purpose of the present invention is to provide a real-time emergency response simulation system and simulation method for a wind farm. The system includes an edge computing module, a central processing module, a quantum computing module, a virtual reality module, an augmented reality module and a biosensor module. The edge computing module is used to detect various real and complex abnormal situations, and the biosensor module is used to adjust the training according to personnel feedback, thereby solving the problems of rapid fault diagnosis of wind farms in complex environments, optimal scheduling of emergency resources under extreme weather conditions, and cognitive load management of operators in high-voltage environments.

[0006] To achieve the above object, the technical solution adopted by the present invention is: A wind farm real-time emergency response simulation system, comprising: An edge computing module, which is disposed on each wind turbine in the wind farm and is used to perform real-time data preprocessing and preliminary fault detection; A central processing module, the central processing module is used to perform in-depth analysis and global decision-making; A virtual reality module, wherein the virtual reality module is used to generate training scenarios and provide virtual training; An augmented reality module, wherein the augmented reality module is used to provide real-time guidance in an actual environment; A quantum computing module, wherein the quantum computing module is used to solve resource scheduling optimization problems; A biosensor module, wherein the biosensor module is used to collect physiological data of an operator; Among them, the edge computing module is communicatively connected with the central processing module, and the central processing module is communicatively connected with the quantum computing module, the virtual reality module, the augmented reality module and the biosensor module respectively; the virtual reality module and the augmented reality module are communicatively connected; the augmented reality module is communicatively connected with the biosensor module; the edge computing module has a vibration data collection device and an acoustic data collection device; the edge computing module has a preliminary fault detection module using an integrated learning method.

[0007] Optionally, the edge computing module has a real-time data preprocessing module, and the real-time data preprocessing module adopts a multi-threaded parallel processing architecture.

[0008] Optionally, the central processing module adopts a large-scale fault diagnosis model of a graph neural network.

[0009] Optionally, the biosensor module has an electrocardiogram (ECG) acquisition device, an electrical skin activity (EGA) acquisition device and an electroencephalogram (EEG) acquisition device.

[0010] The simulation method of the wind farm real-time emergency response simulation system comprises the following steps: Collect wind turbine multi-source data through edge computing module and perform fast anomaly detection in the edge computing module; Transmit abnormal data to the central processing module for in-depth analysis, and send the processing results to the quantum computing module, virtual reality module, augmented reality module and biosensor module; Use quantum computing modules to assist resource optimization and generate long-term resource scheduling strategies; Generate a training environment in the virtual reality module, and send the generated environment to the augmented reality module to train personnel through the augmented reality module; The biosensor module assesses the physiological state of the personnel and dynamically adjusts the training difficulty.

[0011] Optionally, in the step of performing fast anomaly detection through the edge computing module, an integrated learning method is used to perform single anomaly detection.

[0012] Optionally, in the deep analysis step performed by the central processing module, a graph neural network model is used for global fault diagnosis, the model is trained using a small batch gradient descent method, and the loss function uses cross entropy.

[0013] Optionally, in the fast anomaly detection step performed by the edge computing module, a fast Fourier transform algorithm is used on the vibration data and a short-time Fourier transform algorithm is used on the acoustic data.

[0014] Optionally, the quantum computing module adopts a quantum approximate optimization algorithm, and the quantum circuit depth is dynamically adjusted between 10 and 20.

[0015] Optionally, the virtual reality module adopts a scenario based on a conditional generative adversarial network.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The wind farm real-time emergency response simulation system and method provided by the present invention significantly improves the emergency response capability of the wind farm by integrating the edge computing module, the central processing module, the quantum computing module, the virtual reality module, the augmented reality module and the biosensor module. The system of the present invention realizes rapid fault diagnosis in complex environments through the edge computing module, greatly shortens the fault detection and positioning time, and improves the diagnosis accuracy. In terms of resource scheduling, the system can perform intelligent optimization according to the real-time situation through the sub-computing module, improves the resource utilization efficiency, and reduces the emergency response cost. Through virtual reality and augmented reality technology, the system provides operators with a highly realistic training environment and intuitive on-site guidance, which significantly improves the training effect and actual operation ability. At the same time, the biofeedback mechanism of the system of the present invention effectively reduces the error rate of operators in high-voltage environments. The present invention improves the safety and reliability of wind farms, reduces downtime, improves power generation efficiency, and provides technical support for efficient and safe operation of wind farms. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are for explanation purposes only and are not intended to limit the scope of the present invention in any way. In addition, the shapes and proportional dimensions of the components in the drawings are only for illustration purposes to help understand the present invention, and are not intended to specifically limit the shapes and proportional dimensions of the components of the present invention. In the drawings: Figure 1 This is an overall architecture diagram of a wind farm real-time emergency response simulation system according to Embodiment 1 of the present invention; Figure 2 This is a flow chart of a wind farm real-time emergency response simulation method according to Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0019] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] In the description of the embodiments of the present invention, it should be noted that if the terms "upper", "lower", "horizontal", "inner", etc. appear, the orientation or position relationship indicated is based on the orientation or position relationship shown in the accompanying drawings, or is the orientation or position relationship in which the product of the invention is usually placed when used. It is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0021] When an element is referred to as being "disposed on" another element, it may be directly on the other element or there may also be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may also be a central element. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only and are not intended to be the only embodiments. If the term "horizontal" appears, it does not mean that the component is required to be absolutely horizontal, but it can be slightly tilted. For example, "horizontal" only means that its direction is more horizontal than "vertical", which does not mean that the structure must be completely horizontal, but it can be slightly tilted.

[0022] It should be noted that similar reference numerals and letters represent similar items in the following figures, so once an item is defined in one figure, it does not need to be further defined and explained in the subsequent figures. In the description of the present invention, it is to be understood that the terms "include" and "comprise" indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections.

[0023] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art of the technical field of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, the singular forms of "a", "an" and "the" are intended to include plural forms unless the context clearly indicates otherwise.

[0024] The present invention is described in detail below with reference to the accompanying drawings.

[0025] like Figure 1 As shown, a wind farm real-time emergency handling simulation system of the present invention comprises: An edge computing module, which is disposed on each wind turbine in the wind farm and is used to perform real-time data preprocessing and preliminary fault detection; A central processing module, the central processing module is used to perform in-depth analysis and global decision-making; A virtual reality module, wherein the virtual reality module is used to generate training scenarios and provide virtual training; An augmented reality module, wherein the augmented reality module is used to provide real-time guidance in an actual environment; A quantum computing module, wherein the quantum computing module is used to solve resource scheduling optimization problems; A biosensor module, wherein the biosensor module is used to collect physiological data of an operator; Among them, the edge computing module is communicatively connected with the central processing module, and the central processing module is communicatively connected with the quantum computing module, the virtual reality module, the augmented reality module and the biosensor module respectively; the virtual reality module and the augmented reality module are communicatively connected; the augmented reality module is communicatively connected with the biosensor module; the edge computing module has a vibration data collection device and an acoustic data collection device; the edge computing module has a preliminary fault detection module using an integrated learning method.

[0026] The simulation method of the wind farm real-time emergency response simulation system comprises the following steps: Collect wind turbine multi-source data through edge computing module and perform fast anomaly detection in the edge computing module; Transmit abnormal data to the central processing module for in-depth analysis, and send the processing results to the quantum computing module, virtual reality module, augmented reality module and biosensor module; Use quantum computing modules to assist resource optimization and generate long-term resource scheduling strategies; Generate a training environment in the virtual reality module, and send the generated environment to the augmented reality module to train personnel through the augmented reality module; The biosensor module assesses the physiological state of the personnel and dynamically adjusts the training difficulty.

[0027] The present invention provides a wind farm real-time emergency response simulation system and simulation method, which can solve the problems of rapid fault diagnosis of wind farms in complex environments, optimal scheduling of emergency resources under extreme weather conditions, and cognitive load management of operators under high-voltage environments.

[0028] Example 1 This embodiment provides a wind farm real-time emergency response simulation system, and its overall architecture is as follows: Figure 1 shown.

[0029] The system includes: edge computing module, central processing module, quantum computing module, virtual reality module, augmented reality module and biosensor module The edge computing module is set on each wind turbine, adopts a heterogeneous computing architecture, integrates multi-core CPU and parallel GPU, and provides a computing performance of more than 200 TOPS. Its functions include real-time data preprocessing and preliminary fault detection. Specifically, the edge computing module includes a real-time data preprocessing module, a local anomaly detection module, The real-time data preprocessing module uses a multi-threaded parallel processing architecture to simultaneously process high-frequency data streams from different sensors. Specifically, the real-time data preprocessing module includes a vibration data collection device, a temperature data collection device, and an acoustic data collection device.

[0030] For vibration data, the system uses the Fast Fourier Transform (FFT) algorithm, performing a transformation every 256 sampling points with an overlap rate of 50%. The FFT calculation formula is as follows:

[0031] in, is the time domain signal, is the frequency domain result, is the number of sampling points.

[0032] Acoustic data is transformed by short-time Fourier transform (STFT) and a 1024-point Hanning window is used to calculate the power spectrum density. The system introduces adaptive window technology, which is implemented as follows: Signal non-stationarity assessment: Use the sliding variance method to calculate the variance change of the signal within a short time window.

[0033] Window size adjustment: The window size is dynamically adjusted between 512 and 2048 points depending on the degree of non-stationarity.

[0034] Adaptive threshold: Set the initial threshold to 10% of the signal amplitude, update the threshold every 100ms, and use the exponential moving average method to smooth the threshold changes.

[0035] The local anomaly detection module uses an ensemble learning method that combines the Isolation Forest algorithm and the Local Anomaly Factor (LOF) method. The system implements a dynamic feature importance evaluation mechanism. The specific steps are as follows: Initial feature set: 50 key sensor indicators are selected as initial features.

[0036] Periodic evaluation: Feature importance evaluation is performed every hour.

[0037] Evaluation method: Feature importance scoring using random forests combined with mutual information calculation.

[0038] Dynamic adjustment: retain the top 20 features with the highest importance score, and randomly select 5 of the remaining features to retain.

[0039] Model update: Based on the new feature set, the anomaly detection model is updated using an incremental learning method.

[0040] The system also implements a dynamic threshold adjustment mechanism based on statistical process control (SPC): Initial threshold setting: Use the 99.7% quantile of historical data (3σ principle) as the initial threshold.

[0041] Control Chart Construction: Build an Individual Moving Range (IMR) control chart for each key indicator.

[0042] Dynamic adjustment: When 7 consecutive points fall outside the control limits, the control limits are recalculated.

[0043] Smoothing: Use the exponentially weighted moving average (EWMA) method to smooth threshold changes.

[0044] The central processing module is used for in-depth analysis and global decision-making. It deploys a large-scale fault diagnosis model based on graph neural network (GNN). This large-scale fault diagnosis model models the wind farm as a dynamic graph structure, where each wind turbine is a node in the graph and the edges between nodes represent their physical and operational associations.

[0045] The model uses graph convolutional network (GCN) as the basic architecture and introduces gated recurrent module (GRU) to process temporal information. The propagation rule of one layer of GCN can be expressed as:

[0046] in, is the adjacency matrix with self-connection added, yes The degree matrix of It is The node feature matrix of the layer, is a learnable weight matrix, is the activation function.

[0047] The quantum computing module is used to solve complex optimization problems such as resource scheduling. The system uses the quantum approximate optimization algorithm (QAOA), and its quantum circuit depth can be dynamically adjusted between p=10 and 20. The objective function of QAOA can be expressed as:

[0048] in, and is the classic parameter to be optimized, is a quantum circuit, is the cost Hamiltonian.

[0049] The virtual reality module uses scene generation technology based on generative adversarial networks (GAN). The system develops a conditional GAN ​​architecture whose discriminator and generator The objective function is as follows:

[0050] in, It is real scene data. are conditional inputs (such as weather conditions, fault types), is random noise.

[0051] The biosensor module has an electrocardiogram acquisition device, a skin electrode activity acquisition device and an electroencephalogram acquisition device.

[0052] The biosensor module collects real-time physiological data of the operator, including electrocardiogram (ECG), electrodermal activity (EDA), and electroencephalogram (EEG). A multi-scale temporal convolutional network (MS-TCN) is designed to process these multimodal physiological signals. A convolutional block of MS-TCN can be expressed as:

[0053] in, are input features of different scales, is the corresponding convolution kernel, represents the convolution operation, is the bias term.

[0054] Example 2 This embodiment provides a simulation method based on the wind farm real-time emergency response simulation system of embodiment 1. The method flow is as follows: Figure 2 As shown, the steps include: S1: Multi-source data collection and preprocessing At the same time, the vibration, temperature and acoustic data of the fan are collected. For vibration data, the sampling rate is set to 20kHz; the temperature data sampling rate is 1Hz; the acoustic data sampling rate is 44.1kHz.

[0055] S2: Fast anomaly detection at the edge Using the local anomaly detection method in Example 1, fast anomaly detection is performed on the edge computing module.

[0056] S3: In-depth analysis of the central module When the edge module detects an anomaly, the data is transmitted to the central processing module for in-depth analysis. The central module uses the GNN model described in Example 1 for global fault diagnosis. The model training uses the mini-batch gradient descent method, and the loss function uses the cross entropy:

[0057] in, is the true label, is the model prediction value, is the number of samples.

[0058] S4: Quantum-assisted resource optimization For the resource scheduling problem, the system first uses the QAOA algorithm to generate multiple candidate solutions, and then uses the simulated annealing algorithm for further optimization. The parameter optimization of QAOA uses the Adam algorithm, with the learning rate set to 0.01 and the number of iterations set to 1000.

[0059] S5: Reinforcement Learning Policy Generation The improved Proximal Policy Optimization (PPO) algorithm is used to generate long-term resource scheduling strategies. The objective function of PPO is:

[0060] in, is the probability ratio of the new and old strategies, is the advantage function estimate, is the cropping parameter.

[0061] S6: Virtual Reality and Augmented Reality Training and Guidance Training is provided in a virtual reality environment based on generated scenarios and strategies. The difficulty of training is dynamically adjusted based on the operator's performance. In actual operations, real-time guidance is provided through augmented reality devices.

[0062] S7: Biofeedback Assessment The multi-scale temporal convolutional network (MS-TCN) was used to process the physiological data of the operators and estimate their cognitive load level. The network was trained on a labeled dataset containing 500 operators and a total of 1,000 hours, and achieved a 90% accuracy rate in estimating cognitive load within a 5-second time window.

[0063] S8: Adaptive Difficulty Adjustment Based on the estimated cognitive load level, the training difficulty is dynamically adjusted according to the feedback structure of the biosensor module. A personalized difficulty progression curve is generated using a Bayesian optimization algorithm, and the optimization goal is to maximize the learning effect while maintaining the cognitive load at 70%-80% of the estimated maximum cognitive capacity.

[0064] S9: System performance optimization The quantum computing module is used to regularly evaluate overall performance, including fault diagnosis accuracy, resource scheduling efficiency, and operator performance. Based on the evaluation results, the parameters and strategies of each module are updated to continuously optimize system performance.

[0065] Through the above steps, the whole process of intelligent management of wind farm emergency response is realized, which can significantly improve the emergency response speed and decision-making accuracy, and at the same time provide an efficient training and guidance mechanism.

[0066] Unless otherwise specified, the device elements involved in the above embodiments are all conventional device elements, and the structural settings, working modes or control modes involved are all conventional settings, working modes or control modes in the art unless otherwise specified.

[0067] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Other modifications or equivalent substitutions made to the technical solution of the present invention by ordinary technicians in the field should be included in the scope of the claims of the present invention as long as they do not depart from the spirit and scope of the technical solution of the present invention.

Claims

1. A wind farm real-time emergency response simulation system, characterized in that: include: An edge computing module, which is disposed on each wind turbine in the wind farm and is used to perform real-time data preprocessing and preliminary fault detection; A central processing module, the central processing module is used to perform in-depth analysis and global decision-making; A virtual reality module, wherein the virtual reality module is used to generate training scenarios and provide virtual training; An augmented reality module, wherein the augmented reality module is used to provide real-time guidance in an actual environment; A quantum computing module, wherein the quantum computing module is used to solve resource scheduling optimization problems; A biosensor module, wherein the biosensor module is used to collect physiological data of an operator; The edge computing module is connected to the central processing module in communication, and the central processing module is connected to the quantum computing module, the virtual reality module, the augmented reality module and the biosensor module in communication; the virtual reality module is connected to the augmented reality module in communication; the augmented reality module is connected to the biosensor module in communication; the edge computing module has a vibration data collection device and an acoustic data collection device; The edge calculation module has a preliminary fault detection module using an integrated learning method.

2. A wind farm real-time emergency response simulation system according to claim 1, characterized in that: The edge computing module has a real-time data preprocessing module, and the real-time data preprocessing module adopts a multi-threaded parallel processing architecture.

3. A wind farm real-time emergency response simulation system according to claim 1, characterized in that: The central processing module adopts a large-scale fault diagnosis model of a graph neural network.

4. A wind farm real-time emergency response simulation system according to claim 1, characterized in that: The biosensor module has an electrocardiogram acquisition device, a skin electrode activity acquisition device and an electroencephalogram acquisition device.

5. The simulation method of the wind farm real-time emergency handling simulation system according to any one of claims 1 to 4, characterized in that: The following steps are involved: Collect wind turbine multi-source data through edge computing module and perform fast anomaly detection in the edge computing module; Transmit abnormal data to the central processing module for in-depth analysis, and send the processing results to the quantum computing module, virtual reality module, augmented reality module and biosensor module; Use quantum computing modules to assist resource optimization and generate long-term resource scheduling strategies; Generate a training environment in the virtual reality module, and send the generated environment to the augmented reality module to train personnel through the augmented reality module; The biosensor module assesses the physiological state of the personnel and dynamically adjusts the training difficulty.

6. The simulation method of the wind farm real-time emergency handling simulation system according to claim 5, characterized in that: In the step of fast anomaly detection through the edge computing module, an integrated learning method is used for single anomaly detection.

7. The simulation method of the wind farm real-time emergency handling simulation system according to claim 5, characterized in that: In the deep analysis step through the central processing module, a graph neural network model is used for global fault diagnosis, and the model is trained using the small batch gradient descent method, with the cross entropy loss function.

8. The simulation method of the wind farm real-time emergency handling simulation system according to claim 5, characterized in that: In the fast anomaly detection step performed by the edge computing module, the fast Fourier transform algorithm is used for vibration data and the short-time Fourier transform algorithm is used for acoustic data.

9. The simulation method of the wind farm real-time emergency handling simulation system according to claim 5, characterized in that: The quantum computing module adopts a quantum approximate optimization algorithm, and the quantum circuit depth is dynamically adjusted between 10 and 20.

10. The simulation method of the wind farm real-time emergency handling simulation system according to claim 5, characterized in that: The virtual reality module adopts a scenario based on a conditional generative adversarial network.

Images