Self-adaptive control method and system for generator set

Abstract

The invention provides a self-adaptive control method and system for a generator set, and belongs to the technical field of regulation and control, and a controller can carry out the deep analysis and processing of input data through the powerful calculation capability of a neural network model. The neural network model can identify and predict various complex operation modes and potential problems after being trained. The controller can accurately regulate and control the operation process of the generator body in real time according to the output result of the neural network model. The regulation and control not only comprises the adjustment of the power output of the generator, but also comprises the real-time monitoring and optimization of key parameters such as the temperature, the rotating speed and the like of the generator. Through the intelligent regulation and control means, the operation efficiency of the generator can be remarkably improved, the service life of the generator is prolonged, and the stability and reliability of the generator under various working conditions are ensured.

Description

Technical Field

[0001] The present application relates to the field of control technology, and in particular to a method and system for adaptively controlling a generator set. Background Art

[0002] Generator sets usually consist of engines and generators, which can meet various power needs by converting mechanical energy into electrical energy. In order to improve the efficiency and stability of generator sets, adaptive control methods are widely used in their control systems.

[0003] Adaptive control is an intelligent control strategy that can automatically adjust control parameters according to the real-time operating status of the system and changes in the external environment. The core of this method is that it can monitor the performance indicators of the system in real time and optimize the control strategy through specific algorithms to achieve the best control effect.

[0004] Traditional generator control systems mostly use fixed parameter control strategies, which are difficult to cope with complex and changeable working environments and load fluctuations, and the control accuracy is low. Summary of the invention

[0005] The embodiments of the present application provide a method and system for adaptively controlling a generator set to improve the above-mentioned problems.

[0006] In order to achieve the above objectives, this application adopts the following technical solutions:

[0007] In a first aspect, an embodiment of the present application proposes a generator set adaptive control method, which is applicable to a generator set adaptive control system, the system including a generator body, a controller and a sensor, and the method is applicable to the controller, including:

[0008] The controller obtains multiple operating parameters of the generator body during the test cycle based on the sensor, and the multiple operating parameters include: voltage value, current value, frequency value, water temperature value and oil pressure value;

[0009] The controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into the neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and outputs the predicted operating parameters that have not been input according to the combination of any three operating parameters among the input voltage value, the current value, the frequency value, the water temperature value and the oil pressure value;

[0010] Control to obtain the trained neural network model during the operation cycle, and obtain multiple target operation data of the generator body during the test cycle;

[0011] The controller inputs the target operation data into the neural network model, and regulates the operation process of the generator body according to the output results of the neural network model.

[0012] In combination with the first aspect, in some embodiments, the controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into the neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, including:

[0013] The controller uses the voltage value, the current value, the frequency value, and the water temperature value as three different dimensions, which constitute a feature vector of the oil pressure value. A first neural network training model is constructed based on the feature vector of the oil pressure value.

[0014] The controller uses the voltage value, the current value and the oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vectors of the frequency value and the water temperature value. The second neural network training model is constructed based on the characteristic vectors of the frequency value and the water temperature value.

[0015] The controller uses the voltage value, the frequency value, the water temperature value and the oil pressure value as three different dimensions, which constitute the characteristic vector of the current value, and constructs a third neural network training model based on the characteristic vector of the current value;

[0016] The controller takes the current value, frequency value, water temperature value and oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value.

[0017] In combination with the first aspect, in some embodiments, the controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into the neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, including:

[0018] The controller cross-iterates the first neural network model, the second neural network model, the third neural network model and the fourth neural network model, stops iterating when the cross-iteration process meets a preset condition, and selects the corresponding first neural network model, the second neural network model, the third neural network model and the fourth neural network model from the multiple iteration results;

[0019] The corresponding first neural network model, second neural network model, third neural network model and fourth neural network model are selected from multiple iteration results for model fusion to obtain a neural network model.

[0020] In combination with the first aspect, in some implementations, the controller uses the voltage value, the current value, the frequency value, and the water temperature value as three different dimensions, which constitute a feature vector of the oil pressure value. A first neural network training model is constructed based on the feature vector of the oil pressure value, including:

[0021] The controller constructs a first neural network training model based on the characteristic vector of the oil pressure value to perform iterative calculation and obtain a first error value;

[0022] The controller compares the first error value with the first preset value. When the first error value is less than or equal to the first preset value, the controller stops iterating the first neural network prediction model and outputs the iterated first neural network prediction model.

[0023] In combination with the first aspect, in some implementations, the controller uses the voltage value, the current value, and the oil pressure value as three different dimensions, which constitute feature vectors of the frequency value and the water temperature value. A second neural network training model is constructed based on the feature vectors of the frequency value and the water temperature value, including:

[0024] The controller constructs a second neural network training model based on the characteristic vectors of the frequency value and the water temperature value for iterative calculation, and obtains a second error value;

[0025] The controller compares the second error value with the second preset value. When the second error value is less than or equal to the second preset value, the controller stops iterating the second neural network prediction model and outputs the iterated second neural network prediction model.

[0026] In combination with the first aspect, in some implementations, the controller uses the voltage value, the frequency value, the water temperature value, and the oil pressure value as three different dimensions, which constitute a feature vector of the current value, and constructs a third neural network training model based on the feature vector of the current value, including:

[0027] The controller constructs a second neural network training model based on the characteristic vector of the current value for iterative calculation, and obtains a third error value;

[0028] The controller compares the third error value with the third preset value. When the third error value is less than or equal to the third preset value, the controller stops iterating the third neural network prediction model and outputs the iterated third neural network prediction model.

[0029] In combination with the first aspect, in some implementations, the controller uses the current value, the frequency value, the water temperature value, and the oil pressure value as three different dimensions, which constitute a characteristic vector of the voltage value, and constructs a fourth neural network training model based on the characteristic vector of the voltage value, including:

[0030] The controller constructs a fourth neural network training model based on the characteristic vector of the voltage value for iterative calculation, and obtains a fourth error value;

[0031] The controller compares the fourth error value with the fourth preset value. When the fourth error value is less than or equal to the fourth preset value, the controller stops iterating the fourth neural network prediction model and outputs the iterated third neural network prediction model.

[0032] In combination with the first aspect, in some embodiments, the controller cross-iterates the first neural network model, the second neural network model, the third neural network model, and the fourth neural network model, and when the cross-iteration process meets a preset condition, stops the iteration, and selects the corresponding first neural network model, the second neural network model, the third neural network model, and the fourth neural network model from multiple iteration results, including:

[0033] The controller fuses the first error value, the second error value, the third error value, and the fourth error value, and obtains a target demand value according to the fusion result;

[0034] The controller fuses the first neural network model, the second neural network model, the third neural network model and the fourth neural network model according to the target demand value, stops the fusion when the error value of any neural network model among the first neural network model, the second neural network model, the third neural network model and the fourth neural network model is equal to the target demand value, and outputs the neural network model.

[0035] In a second aspect, an embodiment of the present application proposes an adaptive control system for a generator set, the system comprising a generator body, a controller and a sensor, the method is applicable to the controller, and the system is configured as follows:

[0036] The controller obtains multiple operating parameters of the generator body during the test cycle based on the sensor, and the multiple operating parameters include: voltage value, current value, frequency value, water temperature value and oil pressure value;

[0037] The controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into the neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and outputs the predicted operating parameters that have not been input according to the combination of any three operating parameters among the input voltage value, the current value, the frequency value, the water temperature value and the oil pressure value;

[0038] Control to obtain the trained neural network model during the operation cycle, and obtain multiple target operation data of the generator body during the test cycle;

[0039] The controller inputs the target operation data into the neural network model, and regulates the operation process of the generator body according to the output results of the neural network model.

[0040] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0041] The controller imports the voltage value, current value, frequency value, water temperature value and oil pressure value into the neural network model. The neural network model is used to iteratively calculate the voltage value, current value, frequency value, water temperature value and oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, current value, frequency value, water temperature value and oil pressure value, including:

[0042] The controller uses the voltage value, the current value, the frequency value, and the water temperature value as three different dimensions, which constitute a feature vector of the oil pressure value. A first neural network training model is constructed based on the feature vector of the oil pressure value.

[0043] The controller uses the voltage value, the current value and the oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vectors of the frequency value and the water temperature value. The second neural network training model is constructed based on the characteristic vectors of the frequency value and the water temperature value.

[0044] The controller uses the voltage value, the frequency value, the water temperature value and the oil pressure value as three different dimensions, which constitute the characteristic vector of the current value, and constructs a third neural network training model based on the characteristic vector of the current value;

[0045] The controller takes the current value, frequency value, water temperature value and oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value.

[0046] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0047] The controller imports the voltage value, current value, frequency value, water temperature value and oil pressure value into the neural network model. The neural network model is used to iteratively calculate the voltage value, current value, frequency value, water temperature value and oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, current value, frequency value, water temperature value and oil pressure value, including:

[0048] The controller cross-iterates the first neural network model, the second neural network model, the third neural network model and the fourth neural network model, stops iterating when the cross-iteration process meets a preset condition, and selects the corresponding first neural network model, the second neural network model, the third neural network model and the fourth neural network model from the multiple iteration results;

[0049] The corresponding first neural network model, second neural network model, third neural network model and fourth neural network model are selected from multiple iteration results for model fusion to obtain a neural network model.

[0050] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0051] The controller uses the voltage value, current value, frequency value, and water temperature value as three different dimensions, which constitute the characteristic vector of the oil pressure value. The first neural network training model is constructed based on the characteristic vector of the oil pressure value, including:

[0052] The controller constructs a first neural network training model based on the characteristic vector of the oil pressure value to perform iterative calculation and obtain a first error value;

[0053] The controller compares the first error value with the first preset value. When the first error value is less than or equal to the first preset value, the controller stops iterating the first neural network prediction model and outputs the iterated first neural network prediction model.

[0054] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0055] The controller uses the voltage value, current value and oil pressure value as three different dimensions, which constitute the characteristic vectors of the frequency value and the water temperature value. The second neural network training model is constructed based on the characteristic vectors of the frequency value and the water temperature value, including:

[0056] The controller constructs a second neural network training model based on the characteristic vectors of the frequency value and the water temperature value for iterative calculation, and obtains a second error value;

[0057] The controller compares the second error value with the second preset value. When the second error value is less than or equal to the second preset value, the controller stops iterating the second neural network prediction model and outputs the iterated second neural network prediction model.

[0058] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0059] The controller uses the voltage value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the feature vector of the current value. The third neural network training model is constructed based on the feature vector of the current value, including:

[0060] The controller constructs a second neural network training model based on the characteristic vector of the current value for iterative calculation, and obtains a third error value;

[0061] The controller compares the third error value with the third preset value. When the third error value is less than or equal to the third preset value, the controller stops iterating the third neural network prediction model and outputs the iterated third neural network prediction model.

[0062] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0063] The controller uses the current value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value, including:

[0064] The controller constructs a fourth neural network training model based on the characteristic vector of the voltage value for iterative calculation, and obtains a fourth error value;

[0065] The controller compares the fourth error value with the fourth preset value. When the fourth error value is less than or equal to the fourth preset value, the controller stops iterating the fourth neural network prediction model and outputs the iterated third neural network prediction model.

[0066] In conjunction with the second aspect, in some embodiments, the system is configured as follows:

[0067] The controller cross-iterates the first neural network model, the second neural network model, the third neural network model, and the fourth neural network model. When the cross-iteration process meets the preset conditions, the iteration is stopped, and the corresponding first neural network model, the second neural network model, the third neural network model, and the fourth neural network model are selected from the multiple iteration results, including:

[0068] The controller fuses the first error value, the second error value, the third error value, and the fourth error value, and obtains a target demand value according to the fusion result;

[0069] The controller fuses the first neural network model, the second neural network model, the third neural network model and the fourth neural network model according to the target demand value, stops the fusion when the error value of any neural network model among the first neural network model, the second neural network model, the third neural network model and the fourth neural network model is equal to the target demand value, and outputs the neural network model.

[0070] A third aspect of an embodiment of the present invention provides an electronic device, the electronic device comprising:

[0071] At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method proposed in the first aspect of the embodiment of the present invention.

[0072] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method provided in the first aspect of the embodiment of the present invention.

[0073] In summary, the above method and device have the following technical effects:

[0074] An adaptive control method and system for a generator set proposed in an embodiment of the present application, the controller can use the powerful computing power of the neural network model to deeply analyze and process the input data. The neural network model is trained to identify and predict various complex operating modes and potential problems. According to the output results of the neural network model, the controller can accurately control the operation process of the generator body in real time. This regulation not only includes adjusting the power output of the generator, but also includes real-time monitoring and optimization of key parameters such as the temperature and speed of the generator. Through this intelligent control method, the operating efficiency of the generator can be significantly improved, its service life can be extended, and its stability and reliability under various working conditions can be ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 A schematic flow chart of a method for adaptively controlling a generator set proposed in an embodiment of the present application. DETAILED DESCRIPTION

[0076] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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 are within the scope of protection of the present invention.

[0077] See also Figure 1 The embodiment of the present application proposes a generator set adaptive control method, which is applicable to a generator set adaptive control system. The system includes a generator body, a controller and a sensor. The method is applicable to the controller and includes the following steps:

[0078] S101: The controller obtains multiple operating parameters of the generator body during the test cycle based on sensors, and the multiple operating parameters include: voltage value, current value, frequency value, water temperature value and oil pressure value.

[0079] It can be understood that in this step, the controller can monitor and obtain multiple key operating parameters of the generator body in real time during the test cycle based on the data provided by the sensor. These operating parameters specifically include: voltage value, current value, frequency value, water temperature value and oil pressure value. Through the comprehensive analysis of these parameters, the performance and operating status of the generator during the test cycle can be comprehensively evaluated.

[0080] S102: The controller imports the voltage value, current value, frequency value, water temperature value and oil pressure value into the neural network model, and the neural network model is used to iteratively calculate the voltage value, current value, frequency value, water temperature value and oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, current value, frequency value, water temperature value and oil pressure value.

[0081] As you can understand, the controller is responsible for inputting the collected voltage, current, frequency, water temperature and oil pressure values ​​into a pre-trained neural network model. This neural network model is carefully designed to perform complex iterative calculations and analysis on these operating parameters. In this way, the controller can use the powerful computing power of the neural network to perform in-depth processing and analysis on the input voltage, current, frequency, water temperature and oil pressure values.

[0082] Specifically, the neural network model receives these operating parameters as input data and performs complex calculations through its internal multi-layer neurons. These calculation processes include but are not limited to the adjustment of weights, the application of activation functions, and the back propagation of errors. Through these steps, the neural network model can gradually optimize its internal parameters, thereby improving the processing accuracy of the input data.

[0083] For example, the neural network model receives these operating parameters as input data and performs complex calculations through its internal multi-layer neurons. Specifically, each layer of neurons performs weighted summation based on the output of the previous layer, and then performs nonlinear transformation through activation functions. These activation functions can be Sigmoid, ReLU or Tanh, etc., which enable the neural network to capture complex patterns and relationships in the input data.

[0084] After multiple layers of forward propagation, the neural network will generate a prediction result. In order to make this result closer to the true value, the weights and biases in the network need to be adjusted through the back-propagation algorithm. The back-propagation algorithm calculates the gradient of the loss function and transmits the error signal layer by layer to achieve weight optimization.

[0085] During the training process, the neural network will continuously iterate on the training data set, and each iteration will update the parameters in the network. In order to prevent overfitting, regularization techniques such as L1, L2 regularization or Dropout method are usually used. In addition, the adjustment of the learning rate is also a very critical step in the optimization process. Too high or too low a learning rate will affect the convergence speed and final performance of the model.

[0086] After training is completed, the neural network model needs to be evaluated on the validation set and the test set to ensure its generalization ability. By comparing the performance of the model on different data sets, the model structure or parameters can be further adjusted to achieve the best prediction effect. During the training process, the neural network model will learn the complex relationship between voltage values, current values, frequency values, water temperature values, and oil pressure values. This relationship is not limited to the independent influence of a single parameter, but also includes the interaction and combined effects between these parameters. Therefore, when the controller inputs these operating parameters into the neural network model, the model can predict the values ​​of other operating parameters that are not directly input based on the combination of these parameters.

[0087] For example, if the controller only inputs voltage, current, frequency, and water temperature, the neural network model can predict the current system oil pressure based on the combination of these parameters. Similarly, if the controller only inputs current, frequency, water temperature, and oil pressure, the model can also predict the system voltage. This predictive capability enables the controller to have a more comprehensive understanding of the system's operating status, allowing for more effective monitoring and management.

[0088] Specifically, as an implementation method, in this step, the controller takes the voltage value, current value, frequency value and water temperature value as three different dimensions, and the three different dimensions constitute the characteristic vector of the oil pressure value, and constructs the first neural network training model based on the characteristic vector of the oil pressure value.

[0089] The controller takes the voltage value, current value and oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vectors of the frequency value and the water temperature value. The second neural network training model is constructed based on the characteristic vectors of the frequency value and the water temperature value.

[0090] It can be understood that in this step, the controller can construct a first neural network training model with the characteristic vector of the oil pressure value for iterative calculation and obtain a first error value. Then, the controller compares the first error value with a first preset value. When the first error value is less than or equal to the first preset value, the controller stops iterating the first neural network prediction model and outputs the iterated first neural network prediction model.

[0091] Similarly, the controller uses the voltage value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the characteristic vector of the current value, and constructs a third neural network training model based on the characteristic vector of the current value;

[0092] The controller takes the current value, frequency value, water temperature value and oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value.

[0093] It can be understood that the controller constructs a second neural network training model with the characteristic vectors of the frequency value and the water temperature value for iterative calculation and obtains a second error value. Then, the controller compares the second error value with a second preset value. When the second error value is less than or equal to the second preset value, the controller stops iterating the second neural network prediction model and outputs the iterated second neural network prediction model.

[0094] Similarly, the controller uses the voltage value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the characteristic vector of the current value, and constructs a third neural network training model based on the characteristic vector of the current value.

[0095] The controller takes the current value, frequency value, water temperature value and oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value.

[0096] The controller constructs a second neural network training model based on the characteristic vector of the current value for iterative calculation and obtains a third error value. The controller then compares the third error value with a third preset value. When the third error value is less than or equal to the third preset value, the controller stops iterating the third neural network prediction model and outputs the iterated third neural network prediction model.

[0097] Similarly, the controller uses the current value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the characteristic vector of the voltage value, and constructs the fourth neural network training model based on the characteristic vector of the voltage value.

[0098] Then, the controller constructs a fourth neural network training model based on the characteristic vector of the voltage value to perform iterative calculation and obtain a fourth error value;

[0099] The controller compares the fourth error value with the fourth preset value. When the fourth error value is less than or equal to the fourth preset value, the controller stops iterating the fourth neural network prediction model and outputs the iterated third neural network prediction model.

[0100] In this embodiment, the controller cross-iterates the first neural network model, the second neural network model, the third neural network model, and the fourth neural network model. The cross-iteration process involves combining these models with each other and repeatedly training and optimizing them.

[0101] During the cross-iteration process, the controller will continuously monitor whether the iteration process meets certain preset conditions. These preset conditions can be that the number of iterations reaches a certain value, the performance index of the model reaches a predetermined threshold, or the improvement of the model is lower than a minimum value. Once these preset conditions are met, the controller will stop the iteration process.

[0102] After the iteration stops, the controller will select the best model from multiple iteration results. These best models include the first neural network model, the second neural network model, the third neural network model and the fourth neural network model. The selection criteria can be performance indicators such as model accuracy, generalization ability, and computational efficiency.

[0103] Next, the controller will fuse the optimal models selected from multiple iterations. The model fusion process can be a simple weighted average or a more complex fusion strategy, such as stacking, mixing, or ensemble learning. Through model fusion, a neural network model with better overall performance can be obtained.

[0104] Finally, this neural network model with better comprehensive performance will be used as the output result of this embodiment. This model not only integrates the advantages of multiple sub-models, but also can provide higher accuracy and better generalization ability in practical applications, thus performing well in various complex application scenarios.

[0105] Specifically, in this embodiment, the controller fuses the first error value, the second error value, the third error value and the fourth error value, and obtains the target demand value according to the fusion result. Then, the controller fuses the first neural network model, the second neural network model, the third neural network model and the fourth neural network model according to the target demand value. When the error value of any neural network model among the first neural network model, the second neural network model, the third neural network model and the fourth neural network model is equal to the target demand value, the fusion is stopped and the neural network model is output.

[0106] In this system, the controller is responsible for comprehensive processing and fusion of the first error value, the second error value, the third error value and the fourth error value. Through this fusion process, the controller can obtain an accurate target demand value. Subsequently, the controller performs comprehensive fusion processing on the first neural network model, the second neural network model, the third neural network model and the fourth neural network model based on this target demand value. During the fusion process, the controller will continue to monitor the error values ​​of each neural network model. Once it is found that the error value of any one of the first neural network model, the second neural network model, the third neural network model or the fourth neural network model has reached the target demand value, the controller will immediately stop the fusion process and output the neural network model that currently reaches the target demand value as the final result. Such processing ensures that the system can efficiently and accurately select the neural network model that best meets the requirements, thereby improving the performance and reliability of the overall system.

[0107] S103: Control to obtain the trained neural network model during the operation cycle, and obtain multiple target operation data of the generator body during the test cycle.

[0108] It is understandable that during the operation cycle, the controller obtains the trained and optimized neural network model. It can accurately predict and analyze the operating status of the generator. In addition, the control system will also collect multiple target operating data of the generator body during the test cycle. These data include but are not limited to key parameters such as voltage, current, temperature, and speed, which can fully reflect the operating status of the generator in a specific time period. By combining the neural network model and these target operating data, the control system can monitor and evaluate the performance of the generator in real time, thereby ensuring its stable and efficient operation.

[0109] S104: The controller inputs the target operation data into the neural network model, and regulates the operation process of the generator body according to the output results of the neural network model.

[0110] As you can understand, by monitoring the operating status of the generator in real time, the controller can quickly identify potential abnormalities. When the neural network model outputs an abnormal signal, the controller will immediately activate the preset emergency plan. For example, if the temperature of the generator is detected to be too high, the controller will automatically adjust the power of the cooling system to ensure that the equipment operates within a safe temperature range.

[0111] By monitoring the generator's operating status in real time, the controller can quickly identify potential abnormalities. This monitoring system can detect any deviations from the normal range by continuously tracking the generator's key parameters, such as voltage, current, temperature and vibration. The controller uses advanced algorithms and data analysis technology to process and analyze this data in real time, thereby quickly identifying small changes that may indicate equipment failure or performance degradation. This timely identification and response mechanism can not only effectively prevent sudden equipment failures, but also extend the service life of the generator and ensure the stable operation of the entire system.

[0112] By monitoring the generator's operating status in real time, the controller can quickly identify potential abnormalities. Specifically, it analyzes key parameters such as the generator's temperature, voltage, current and frequency. Once a parameter is detected to deviate from the normal range, the controller will immediately issue an alarm and automatically adjust the generator's operating status according to a preset program to avoid further damage.

[0113] In addition, the controller has a self-diagnostic function that can regularly check whether its sensors and actuators are working properly. If any fault is found, the controller will record a detailed error log and send the information to maintenance personnel through the remote communication module. In this way, maintenance personnel can intervene before the problem becomes serious, thus ensuring the continued stable operation of the generator.

[0114] To further improve the reliability of the system, the controller also supports a variety of redundant designs. For example, it is equipped with dual power modules. Even if one module fails, the other module can immediately take over the work to ensure that the controller will not shut down due to power problems. At the same time, the controller also has a hot-swap function, allowing certain key components to be replaced without stopping, thereby minimizing maintenance time.

[0115] In actual applications, the controller can also be linked with other system equipment. For example, it can work with the power distribution system to automatically adjust the output power of the generator according to changes in the grid load. In addition, the controller also supports docking with the intelligent monitoring platform, allowing managers to view the operating status of the generator in real time through mobile phones or computers, and remotely control and adjust it. In addition, the controller can also optimize the maintenance cycle of the generator based on the output results of the neural network model. By analyzing historical operating data and current operating status, the neural network model can predict the degree of wear and potential failure points of the equipment. The controller will arrange a reasonable maintenance plan based on these prediction results, thereby reducing unnecessary downtime and improving power generation efficiency.

[0116] An adaptive control method for a generator set proposed in an embodiment of the present application, the controller can use the powerful computing power of the neural network model to deeply analyze and process the input data. The neural network model is trained to identify and predict various complex operating modes and potential problems. According to the output results of the neural network model, the controller can accurately control the operation process of the generator body in real time. This regulation not only includes adjusting the power output of the generator, but also includes real-time monitoring and optimization of key parameters such as the temperature and speed of the generator. Through this intelligent control method, the operating efficiency of the generator can be significantly improved, its service life can be extended, and its stability and reliability under various working conditions can be ensured.

[0117] Based on the same inventive concept, the embodiment of the present application proposes an adaptive control system for a generator set, the system includes a generator body, a controller and a sensor, the method is applicable to the controller, and the system is configured as follows:

[0118] The controller obtains multiple operating parameters of the generator body during the test cycle based on the sensor, and the multiple operating parameters include: voltage value, current value, frequency value, water temperature value and oil pressure value;

[0119] The controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into the neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and outputs the predicted operating parameters that have not been input according to the combination of any three operating parameters among the input voltage value, the current value, the frequency value, the water temperature value and the oil pressure value;

[0120] Control to obtain the trained neural network model during the operation cycle, and obtain multiple target operation data of the generator body during the test cycle;

[0121] The controller inputs the target operation data into the neural network model, and regulates the operation process of the generator body according to the output results of the neural network model.

[0122] In some embodiments, the system is configured to:

[0123] The controller imports the voltage value, current value, frequency value, water temperature value and oil pressure value into the neural network model. The neural network model is used to iteratively calculate the voltage value, current value, frequency value, water temperature value and oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, current value, frequency value, water temperature value and oil pressure value, including:

[0124] The controller uses the voltage value, the current value, the frequency value, and the water temperature value as three different dimensions, which constitute a feature vector of the oil pressure value. A first neural network training model is constructed based on the feature vector of the oil pressure value.

[0125] The controller uses the voltage value, the current value and the oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vectors of the frequency value and the water temperature value. The second neural network training model is constructed based on the characteristic vectors of the frequency value and the water temperature value.

[0126] The controller uses the voltage value, the frequency value, the water temperature value and the oil pressure value as three different dimensions, which constitute the characteristic vector of the current value, and constructs a third neural network training model based on the characteristic vector of the current value;

[0127] The controller takes the current value, frequency value, water temperature value and oil pressure value as three different dimensions, and the three different dimensions constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value.

[0128] In some embodiments, the system is configured to:

[0129] The controller imports the voltage value, current value, frequency value, water temperature value and oil pressure value into the neural network model. The neural network model is used to iteratively calculate the voltage value, current value, frequency value, water temperature value and oil pressure value, and output the predicted uninput operating parameters according to the combination of any three operating parameters among the input voltage value, current value, frequency value, water temperature value and oil pressure value, including:

[0130] The controller cross-iterates the first neural network model, the second neural network model, the third neural network model and the fourth neural network model, stops iterating when the cross-iteration process meets a preset condition, and selects the corresponding first neural network model, the second neural network model, the third neural network model and the fourth neural network model from the multiple iteration results;

[0131] The corresponding first neural network model, second neural network model, third neural network model and fourth neural network model are selected from multiple iteration results for model fusion to obtain a neural network model.

[0132] In some embodiments, the system is configured to:

[0133] The controller uses the voltage value, current value, frequency value, and water temperature value as three different dimensions, which constitute the characteristic vector of the oil pressure value. The first neural network training model is constructed based on the characteristic vector of the oil pressure value, including:

[0134] The controller constructs a first neural network training model based on the characteristic vector of the oil pressure value to perform iterative calculation and obtain a first error value;

[0135] The controller compares the first error value with the first preset value. When the first error value is less than or equal to the first preset value, the controller stops iterating the first neural network prediction model and outputs the iterated first neural network prediction model.

[0136] In some embodiments, the system is configured to:

[0137] The controller uses the voltage value, current value and oil pressure value as three different dimensions, which constitute the characteristic vectors of the frequency value and the water temperature value. The second neural network training model is constructed based on the characteristic vectors of the frequency value and the water temperature value, including:

[0138] The controller constructs a second neural network training model based on the characteristic vectors of the frequency value and the water temperature value for iterative calculation, and obtains a second error value;

[0139] The controller compares the second error value with the second preset value. When the second error value is less than or equal to the second preset value, the controller stops iterating the second neural network prediction model and outputs the iterated second neural network prediction model.

[0140] In some embodiments, the system is configured to:

[0141] The controller uses the voltage value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the feature vector of the current value. The third neural network training model is constructed based on the feature vector of the current value, including:

[0142] The controller constructs a second neural network training model based on the characteristic vector of the current value for iterative calculation, and obtains a third error value;

[0143] The controller compares the third error value with the third preset value. When the third error value is less than or equal to the third preset value, the controller stops iterating the third neural network prediction model and outputs the iterated third neural network prediction model.

[0144] In some embodiments, the system is configured to:

[0145] The controller uses the current value, frequency value, water temperature value and oil pressure value as three different dimensions, which constitute the characteristic vector of the voltage value. The fourth neural network training model is constructed based on the characteristic vector of the voltage value, including:

[0146] The controller constructs a fourth neural network training model based on the characteristic vector of the voltage value for iterative calculation, and obtains a fourth error value;

[0147] The controller compares the fourth error value with the fourth preset value. When the fourth error value is less than or equal to the fourth preset value, the controller stops iterating the fourth neural network prediction model and outputs the iterated third neural network prediction model.

[0148] In some embodiments, the system is configured to:

[0149] The controller cross-iterates the first neural network model, the second neural network model, the third neural network model, and the fourth neural network model. When the cross-iteration process meets the preset conditions, the iteration is stopped, and the corresponding first neural network model, the second neural network model, the third neural network model, and the fourth neural network model are selected from the multiple iteration results, including:

[0150] The controller fuses the first error value, the second error value, the third error value, and the fourth error value, and obtains a target demand value according to the fusion result;

[0151] The controller fuses the first neural network model, the second neural network model, the third neural network model and the fourth neural network model according to the target demand value, stops the fusion when the error value of any neural network model among the first neural network model, the second neural network model, the third neural network model and the fourth neural network model is equal to the target demand value, and outputs the neural network model.

[0152] An adaptive control system for a generator set proposed in an embodiment of the present application, the controller can use the powerful computing power of the neural network model to deeply analyze and process the input data. The neural network model is trained to identify and predict various complex operating modes and potential problems. According to the output results of the neural network model, the controller can accurately control the operation process of the generator body in real time. This regulation not only includes adjusting the power output of the generator, but also includes real-time monitoring and optimization of key parameters such as the temperature and speed of the generator. Through this intelligent control method, the operating efficiency of the generator can be significantly improved, its service life can be extended, and its stability and reliability under various working conditions can be ensured.

[0153] Based on the same inventive concept, an embodiment of the present application further proposes an electronic device, the electronic device comprising:

[0154] At least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the generator set adaptive control method of the embodiment of the present application.

[0155] In addition, to achieve the above-mentioned purpose, an embodiment of the present application further proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the generator set adaptive control method of the embodiment of the present application.

[0156] The following is a detailed introduction to the various components of electronic equipment:

[0157] The processor is the control center of the electronic device, which can be a processor or a general term for multiple processing elements. For example, the processor is one or more central processing units (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention, such as one or more microprocessors (digital signal processors, DSP), or one or more field programmable gate arrays (field programmable gate arrays, FPGA).

[0158] Optionally, the processor can perform various functions of the electronic device by running or executing a software program stored in the memory, and calling data stored in the memory.

[0159] The memory is used to store the software program for executing the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can refer to the above method embodiment and will not be repeated here.

[0160] Optionally, the memory may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may be integrated with the processor, or may exist independently and be coupled to the processor through an interface circuit of the electronic device, and the embodiments of the present invention do not specifically limit this.

[0161] A transceiver is used to communicate with a network device or a terminal device.

[0162] Optionally, the transceiver may include a receiver and a transmitter, wherein the receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.

[0163] Optionally, the transceiver may be integrated with the processor, or may exist independently and be coupled to the processor via an interface circuit of the router, which is not specifically limited in the embodiment of the present invention.

[0164] In addition, the technical effects of the electronic device can refer to the technical effects of the data transmission method in the above method embodiment, which will not be repeated here.

[0165] It should be understood that the processor in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0166] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus RAM (DRRAM).

[0167] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When a computer instruction or computer program is loaded or executed on a computer, a process or function according to an embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0168] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0169] In the present invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0170] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0171] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

Claims

1. A method for adaptive control of a generator set, characterized in that: The invention is applicable to a generator set adaptive control system, the system comprising a generator body, a controller and a sensor, the method is applicable to the controller, and comprises: The controller acquires a plurality of operating parameters of the generator body during a test cycle based on the sensor, the plurality of operating parameters including: voltage value, current value, frequency value, water temperature value and oil pressure value; The controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into a neural network model, the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and output the predicted operating parameters that have not been input according to a combination of any three of the input operating parameters among the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value; The control acquires the trained neural network model during the operation cycle, and acquires a plurality of target operation data of the generator body during the test cycle; The controller inputs the target operation data into the neural network model, and regulates the operation process of the generator body according to the output result of the neural network model.

2. A method for adaptive control of a generator set according to claim 1, characterized in that: The controller imports the voltage value, the current value, the frequency value, the water temperature value, and the oil pressure value into a neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value, and the oil pressure value, and output the predicted operating parameters that have not been input according to a combination of any three of the input operating parameters among the voltage value, the current value, the frequency value, the water temperature value, and the oil pressure value, including: The controller uses the voltage value, the current value, the frequency value, and the water temperature value as three different dimensions, which constitute a feature vector of the oil pressure value, and constructs a first neural network training model based on the feature vector of the oil pressure value; The controller uses the voltage value, the current value, and the oil pressure value as three different dimensions, which constitute the feature vectors of the frequency value and the water temperature value, and constructs a second neural network training model based on the feature vectors of the frequency value and the water temperature value; The controller uses the voltage value, the frequency value, the water temperature value, and the oil pressure value as three different dimensions, which constitute a feature vector of the current value, and constructs a third neural network training model based on the feature vector of the current value; The controller uses the current value, the frequency value, the water temperature value and the oil pressure value as three different dimensions, which constitute a feature vector of the voltage value, and constructs a fourth neural network training model based on the feature vector of the voltage value.

3. A method for adaptive control of a generator set according to claim 2, characterized in that: The controller imports the voltage value, the current value, the frequency value, the water temperature value, and the oil pressure value into a neural network model, and the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value, and the oil pressure value, and output the predicted operating parameters that have not been input according to a combination of any three of the input operating parameters among the voltage value, the current value, the frequency value, the water temperature value, and the oil pressure value, including: The controller cross-iterates the first neural network model, the second neural network model, the third neural network model and the fourth neural network model, stops iterating when the cross-iteration process meets a preset condition, and selects the corresponding first neural network model, the second neural network model, the third neural network model and the fourth neural network model from multiple iteration results; The corresponding first neural network model, the second neural network model, the third neural network model and the fourth neural network model are selected from the multiple iteration results for model fusion to obtain the neural network model.

4. A method for adaptive control of a generator set according to claim 3, characterized in that: The controller uses the voltage value, the current value, the frequency value, and the water temperature value as three different dimensions, which constitute a feature vector of the oil pressure value. A first neural network training model is constructed based on the feature vector of the oil pressure value, including: The controller constructs a first neural network training model based on the characteristic vector of the oil pressure value to perform iterative calculation and obtain a first error value; The controller compares the first error value with a first preset value. When the first error value is less than or equal to the first preset value, the controller stops iterating the first neural network prediction model and outputs the iterated first neural network prediction model.

5. A method for adaptive control of a generator set according to claim 4, characterized in that: The controller uses the voltage value, the current value, and the oil pressure value as three different dimensions, which constitute the feature vectors of the frequency value and the water temperature value. A second neural network training model is constructed based on the feature vectors of the frequency value and the water temperature value, including: The controller constructs a second neural network training model based on the characteristic vectors of the frequency value and the water temperature value to perform iterative calculation and obtain a second error value; The controller compares the second error value with a second preset value. When the second error value is less than or equal to the second preset value, the controller stops iterating the second neural network prediction model and outputs the iterated second neural network prediction model.

6. A method for adaptive control of a generator set according to claim 5, characterized in that: The controller uses the voltage value, the frequency value, the water temperature value, and the oil pressure value as three different dimensions, which constitute a feature vector of the current value, and constructs a third neural network training model based on the feature vector of the current value, including: The controller constructs a second neural network training model based on the characteristic vector of the current value to perform iterative calculation and obtain a third error value; The controller compares the third error value with a third preset value. When the third error value is less than or equal to the third preset value, the controller stops iterating the third neural network prediction model and outputs the iterated third neural network prediction model.

7. A method for adaptive control of a generator set according to claim 6, characterized in that: The controller uses the current value, the frequency value, the water temperature value, and the oil pressure value as three different dimensions, which constitute a feature vector of the voltage value. A fourth neural network training model is constructed based on the feature vector of the voltage value, including: The controller constructs a fourth neural network training model based on the characteristic vector of the voltage value to perform iterative calculation and obtain a fourth error value; The controller compares the fourth error value with a fourth preset value. When the fourth error value is less than or equal to the fourth preset value, the controller stops iterating the fourth neural network prediction model and outputs the iterated third neural network prediction model.

8. A method for adaptive control of a generator set according to claim 7, characterized in that: The controller cross-iterates the first neural network model, the second neural network model, the third neural network model, and the fourth neural network model, stops iterating when the cross-iteration process meets a preset condition, and selects the corresponding first neural network model, the second neural network model, the third neural network model, and the fourth neural network model from multiple iteration results, including: The controller fuses the first error value, the second error value, the third error value, and the fourth error value, and obtains a target demand value according to a fusion result; The controller fuses the first neural network model, the second neural network model, the third neural network model and the fourth neural network model according to the target demand value, stops fusing when the error value of any neural network model among the first neural network model, the second neural network model, the third neural network model and the fourth neural network model is equal to the target demand value, and outputs the neural network model.

9. A generator set adaptive control system, characterized in that: The system includes a generator body, a controller and a sensor, the method is applicable to the controller, and the system is configured as follows: The controller acquires a plurality of operating parameters of the generator body during a test cycle based on the sensor, the plurality of operating parameters including: voltage value, current value, frequency value, water temperature value and oil pressure value; The controller imports the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value into a neural network model, the neural network model is used to iteratively calculate the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value, and output the predicted operating parameters that have not been input according to a combination of any three of the input operating parameters among the voltage value, the current value, the frequency value, the water temperature value and the oil pressure value; The control acquires the trained neural network model during the operation cycle, and acquires a plurality of target operation data of the generator body during the test cycle; The controller inputs the target operation data into the neural network model, and regulates the operation process of the generator body according to the output result of the neural network model.

10. An electronic device, characterized in that: Electronic equipment includes: at least one processor; And, a memory communicatively connected to at least one of the processors; wherein the memory stores instructions executable by at least one of the processors, and the instructions are executed by at least one of the processors so that at least one of the processors can execute the method as set forth in any one of claims 1-8.

Images