Real-time monitoring system and method of plasma generator

Abstract

The invention relates to the technical field of ion generators, and discloses a real-time monitoring system and method for a plasma generator, and the method comprises the steps: collecting and preprocessing the historical operation data, environmental conditions and configuration parameters of the plasma generator, and constructing an operation index prediction model. The method comprises the following steps: acquiring current data in real time and inputting the data into a model, predicting a future operation index, comparing with a real-time operation index, and marking a deviation point; a state monitoring process is modeled as a Markov decision process, a state space, an action space, a reward function and an output strategy are defined, and actions are optimized and adjusted through a reinforcement learning algorithm. Real-time monitoring is carried out based on the trained state monitoring decision model, and the optimal adjustment action is executed automatically or after confirmation of an operator. According to the invention, intelligent state monitoring and predictive maintenance of the plasma generator are realized, the equipment performance and the system stability are effectively improved, and the industrial application value is remarkable.

Description

Technical Field

[0001] The invention relates to the technical field of ion generators, and in particular to a real-time monitoring system and method for a plasma generator. Background Art

[0002] In industrial production, scientific research experiments and many high-tech fields, plasma generators are an important equipment, and their performance and stability have a vital impact on the operating efficiency of the entire system and the accuracy of the results. However, during the operation of plasma generators, they are affected by many factors, such as changes in environmental conditions, aging of the equipment itself, and adjustment of configuration parameters, and their operating indicators often show complex and changeable trends.

[0003] With the advent of the Industrial 4.0 era and the development of intelligent technology, real-time monitoring and predictive maintenance of plasma generators have become an urgent need in the industry. Although existing monitoring technologies have achieved real-time monitoring of the operating status of equipment to a certain extent, most of them only stay at the level of data collection and simple alarms, lacking in-depth data analysis and prediction capabilities, and cannot accurately predict the future operating status of the equipment, let alone actively adjust equipment parameters or take maintenance measures in advance based on the prediction results.

[0004] Due to the complex working principle of the plasma generator, its performance is affected by multiple factors. How to comprehensively consider environmental conditions, equipment configuration and operation history, build a model that can accurately predict the changing trend of equipment operation indicators, and realize intelligent status monitoring and decision-making on this basis is a problem that current technology needs to solve. Summary of the invention

[0005] The object of the present invention is to provide a real-time monitoring system and method for a plasma generator to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a real-time monitoring method for a plasma generator, the method comprising:

[0007] Collect the time series data of the operating indicators of the plasma generator in historical operation, as well as the environmental conditions during operation and the configuration parameters of the plasma generator; and pre-process the collected data to form a feature data set;

[0008] A prediction model for the operation index of the plasma generator is constructed. The model is based on the characteristic data set, combined with environmental conditions and plasma generator configuration parameters, to predict the change trend of the operation index of the plasma generator in the future.

[0009] Acquire the current configuration parameter data and environmental condition data of the plasma generator in real time, input the acquired data into the operation index prediction model, and receive the predicted operation index output by the model;

[0010] Acquire real-time operating index data of the plasma generator; plot the acquired real-time operating index data and the predicted operating index data output by the model into time-varying curves, align the timestamps of the two curves, and place the two curves in the same coordinate system for comparison; further set a performance deviation threshold to determine whether each corresponding point on the two curves has a deviation exceeding the threshold, and if so, mark it as a deviation point;

[0011] A state monitoring decision model is constructed, and the state monitoring process of the plasma generator is modeled as a Markov decision process. The specific definitions include: ① State space: defined as the combination of real-time operating indicator data, configuration parameters and environmental condition data of the plasma generator; ② Action space: defined as a set of executable monitoring or adjustment operations; ③ Reward function: designed according to the stable operation and performance optimization goals of the plasma generator, including reward mechanisms for state stability, performance maintenance and energy efficiency improvement. At the same time, combined with the monitored performance deviation data, additional rewards are given to operations that reduce deviations, and additional penalties are given to operations that increase deviations; ④ Output strategy: defined as a method for selecting the optimal action under the current state, and the generated adjustment actions are continuously optimized through reinforcement learning algorithms;

[0012] Real-time monitoring is performed based on the trained state monitoring decision model. The configuration parameters and environmental condition data of the plasma generator are collected in real time as model input. The optimal adjustment action is selected according to the current state and strategy, and the decision result is automatically executed or transmitted to the operator through the monitoring system for confirmation before execution.

[0013] Preferably, the operating indicators include plasma density, temperature, voltage, current, power, gas flow and cooling system status; the environmental conditions include ambient temperature, humidity, air pressure and electromagnetic interference; and the configuration parameters include the type, size, operating frequency and electrode material of the plasma generator.

[0014] Preferably, the executable monitoring or adjustment operation set includes: adjusting input power or operating frequency, changing gas type or flow, starting or stopping the cooling system, adjusting electrode spacing or replacing electrodes, cleaning or replacing components, adjusting control system parameters, adjusting data acquisition frequency or monitoring accuracy.

[0015] Preferably, the reward function R is composed of multiple sub-reward items, and the total reward value is obtained by weighted summation. The specific formula is: R = w1*R_stab+w2*R_perf+w3*R_eff+w4*R_maint+w5*R_dev; wherein R_stab represents the reward item for state stability, which is calculated according to the degree of fluctuation of the operating state of the plasma generator; R_perf represents the reward item for performance maintenance, which is calculated according to the maintenance of the key performance indicators of the plasma generator; R_eff represents the reward item for energy efficiency improvement. The reward item is calculated based on the degree of improvement in energy consumption efficiency; R_maint represents the reward item for preventive maintenance, which is calculated based on the execution and effect of preventive maintenance operations; R_dev represents the reward item or penalty item for reducing performance deviation. According to the monitored performance deviation data, positive rewards are given to operations that reduce deviations, and penalties are given to operations that increase deviations. Its value is set to a function that is inversely proportional or directly proportional to the change in deviation; w1, w2, w3, w4 and w5 are the weight coefficients of each sub-reward item, and w1+w2+w3+w4+w5=1.

[0016] Preferably, the operation index prediction model is constructed using a gated recurrent unit (GRU) algorithm, and its network structure specifically includes:

[0017] Input layer: used to receive feature data sets as input, wherein the feature data sets include time series performance indicator data, environmental parameter data, system configuration data, and time-related features;

[0018] Hidden layer: contains at least one GRU hidden layer. Each GRU hidden layer contains several GRU units. Each GRU unit has an update gate and a reset gate to control the flow of information and state updates.

[0019] Output layer: Receives the output of the hidden layer and maps the output of the GRU network to the predicted performance indicator value through a fully connected layer or regression layer.

[0020] Preferably, the step of training the operation indicator prediction model includes:

[0021] S1: Initialize the weights and biases of the GRU network using small random numbers;

[0022] S2: Select the loss function to measure the difference between the model prediction value and the true value;

[0023] S3: Use the back propagation algorithm and gradient descent optimization algorithm to calculate the gradient according to the loss function, update the weights and bias items of the GRU network, and iteratively train the network until the loss function converges;

[0024] S4: Save the trained GRU network model as an operation indicator prediction model in the real-time monitoring method of the plasma generator.

[0025] Preferably, a deep Q network DQN algorithm is used to train the state monitoring decision model.

[0026] Preferably, the step of training the state monitoring decision model includes:

[0027] A1. Initialize the Q network: build a deep neural network to evaluate the value of state-action pairs;

[0028] A2. Initialize the experience playback buffer D: Set the buffer size N to store the state transition experience generated during the plasma generator monitoring process, including state s, action a, reward r, next state s' and whether the termination flag is done;

[0029] A3. For each training round:

[0030] A301: Initialize the state s at the beginning of the round;

[0031] A302: Before the round ends, the following steps are executed in a loop: i) The Q network selects action a according to the current state s, and adopts the ε-greedy strategy to balance exploration and utilization; ii) Execute action a, observe the reward r and the next state s', and whether the termination flag is done; iii) Store the experience (s, a, r, s', done) in the experience replay buffer D; iv) Randomly sample a batch of experience from the experience replay buffer D to update the network parameters; v) Use the sampled experience to update the parameters of the Q network to minimize the prediction error of the value function; vi) Update the current state s to the next state s';

[0032] A303: When the round ends, reset state s to the new round start state;

[0033] A4: During the training process, the performance of the state monitoring decision model is regularly evaluated, and the weight coefficient of the reward function, the learning rate of the model, and the size of the exploration noise are adjusted based on the evaluation results;

[0034] A5: Repeat steps A3 and A4 until a satisfactory performance level is achieved.

[0035] Preferably, in step A4, mean square error (MSE) is used as the performance evaluation index of the current round of training, and the specific implementation steps include:

[0036] A401: Based on the sampled experience tuple (s, a, r, s'), the target Q value of each experience tuple is calculated using the target Q network;

[0037] A402: Calculate the predicted Q value of the main Q network for each state-action pair (s, a) in the experience tuple;

[0038] A403: Calculate the mean square error (MSE) between the predicted Q value and the target Q value as the loss function value of the current iteration. The calculation formula of MSE is:

[0039]

[0040] Among them, N is the number of sampled experience tuples, Q target is the target Q value, Q predicted is the predicted Q value of the main Q network;

[0041] A404: Based on the calculated MSE value, the parameters of the main Q network are updated using the gradient descent algorithm to minimize the MSE value, thereby optimizing the performance of the main Q network;

[0042] A405: Repeat steps A401 to A404 until the end condition of the current training round is met, and then enter the next training round.

[0043] Preferably, a real-time monitoring system for a plasma generator comprises:

[0044] A data collection module is used to collect the time series data of the operating indicators of the plasma generator in the historical operation, the environmental conditions during the operation and the configuration parameters of the plasma generator, and pre-process the collected data to form a feature data set;

[0045] A prediction model building module is used to build an operation index prediction model based on a feature data set, combined with environmental conditions and plasma generator configuration parameters, so as to predict the change trend of the operation index of the plasma generator in the future;

[0046] The prediction data acquisition module is used to obtain the current configuration parameter data and environmental condition data of the plasma generator in real time, and input the acquired data into the operation index prediction module, and receive and output the operation index predicted by the model;

[0047] A real-time data acquisition module is used to obtain real-time operating index data of the plasma generator;

[0048] Deviation detection module, used to draw the real-time operation index data and the predicted operation index data into time-varying curves, align the timestamps and place them in the same coordinate system for comparison, further set the performance deviation threshold, determine whether each corresponding point on the two curves has a deviation exceeding the threshold, and mark the deviation point;

[0049] The state monitoring decision module is used to build a state monitoring decision model, modeling the state monitoring process of the plasma generator as a Markov decision process, including defining the state space, action space, reward function and output strategy, and continuously optimizing the generated adjustment actions through the reinforcement learning algorithm;

[0050] The decision execution module is used to perform real-time monitoring based on the trained state monitoring decision model, collect the configuration parameters and environmental condition data of the plasma generator in real time as model input, select the optimal adjustment action according to the current state and strategy, and automatically execute the decision result or transmit it to the operator for confirmation through the monitoring system.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The present invention collects and analyzes the operating data of the plasma generator in real time, combines environmental conditions and configuration parameters, and constructs an operating index prediction model. The model can accurately predict the changing trend of the operating index of the equipment in the future, so that the operator can understand the equipment status in advance, adjust parameters or take maintenance measures in time, thereby significantly improving the operating efficiency and stability of the plasma generator. Through real-time monitoring and predictive maintenance, the present invention can timely discover and solve potential problems, avoid serious damage to the equipment due to failure, and effectively extend the service life of the plasma generator. At the same time, based on the monitoring of performance deviations and the optimization adjustment of the Markov decision process, it can ensure that the equipment is always in the best working state and reduce wear and aging.

[0053] The present invention applies a reinforcement learning algorithm to the state monitoring decision model of a plasma generator, and realizes intelligent state monitoring and decision-making by continuously optimizing the generated adjustment actions. This intelligent monitoring method not only improves the accuracy and efficiency of monitoring, but also reduces the dependence on manual intervention, and improves the intelligence level of the entire system. The reward function design in the state monitoring decision model fully considers the stable operation, performance maintenance and energy efficiency improvement goals of the plasma generator. By continuously optimizing the adjustment actions, the present invention can achieve performance optimization and energy efficiency improvement while ensuring the stable operation of the equipment, reduce energy consumption, and improve production efficiency.

[0054] The real-time monitoring system and method of the present invention have good maintainability and scalability. The system can automatically record the operation history of the equipment and provide data support for subsequent maintenance and analysis. The system has a clear structure and reasonable module division, which is convenient for subsequent upgrades and expansions to meet the monitoring needs of plasma generators of different models and specifications. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A step diagram of the real-time monitoring method of a plasma generator according to the present invention;

[0056] Figure 2 A training flowchart for the running indicator prediction model;

[0057] Figure 3 Flowchart for building a condition monitoring decision model. DETAILED DESCRIPTION

[0058] 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 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 are within the scope of protection of the present invention.

[0059] See also Figure 1-3 The present invention provides a real-time monitoring method for a plasma generator, the method comprising:

[0060] The operation indicator time series data is collected from the historical operation of the plasma generator, including but not limited to key indicators such as voltage, current, and temperature. At the same time, the environmental conditions during operation (such as ambient temperature, humidity, etc.) and the configuration parameters of the plasma generator (such as operating frequency, power setting, etc.) are recorded. The collected data is preprocessed through cleaning, denoising, and normalization to form a feature data set, which provides a basis for subsequent modeling.

[0061] Based on the feature data set, a machine learning algorithm is used to build an operation index prediction model. The model comprehensively considers environmental conditions and plasma generator configuration parameters, and learns the complex relationship between them through training, so as to predict the change trend of the plasma generator's operation indicators in the future.

[0062] The current configuration parameter data and environmental condition data of the plasma generator are obtained in real time as model input, and the predicted operating indicators output by the model are received. At the same time, the real-time operating indicator data of the plasma generator is obtained for comparison with the predicted results.

[0063] The real-time operation index data and the predicted operation index data are plotted as time-varying curves, and the timestamps are aligned. The two curves are compared in the same coordinate system, and the performance deviation threshold is set to automatically determine whether each corresponding point has a deviation exceeding the threshold. If so, it is marked as a deviation point as the basis for subsequent decision-making.

[0064] Construction of condition monitoring decision model: The condition monitoring process of the plasma generator is modeled as a Markov decision process. The specific definitions include:

[0065] ① State space: It is composed of the real-time operating index data, configuration parameters and environmental condition data of the plasma generator.

[0066] ②Action space: Contains a set of executable monitoring or adjustment operations, such as adjusting operating frequency, power, etc.

[0067] ③ Reward function: Designed according to the stable operation and performance optimization goals of the plasma generator, it includes reward mechanisms for state stability, performance maintenance, and energy efficiency improvement. At the same time, combined with the monitored performance deviation data, additional rewards are given to operations that reduce deviations, and additional penalties are given to operations that increase deviations.

[0068] ④Output strategy: Use reinforcement learning algorithm to continuously optimize the generated adjustment actions and select the optimal action under the current state.

[0069] Based on the trained state monitoring decision model, the configuration parameters and environmental condition data of the plasma generator are collected in real time as model input. According to the current state and strategy, the optimal adjustment action is selected. The decision result can be executed automatically or transmitted to the operator through the monitoring system for confirmation and execution, ensuring the stable operation and performance optimization of the plasma generator.

[0070] The present invention will be further described below in conjunction with Examples 1 to 4:

[0071] Embodiment 1:

[0072] In the selection of operating indicators, the present invention comprehensively considers the key performance parameters of the plasma generator, including but not limited to plasma density, temperature, voltage, current, power, gas flow rate and cooling system status. These indicators can fully reflect the working state and performance level of the plasma generator. Plasma density and temperature are important parameters for describing the plasma state, which directly affect the generation and maintenance of plasma. Voltage, current and power reflect the working efficiency and energy consumption of the plasma generator. Gas flow rate determines the generation rate and stability of plasma. The cooling system status is related to whether the plasma generator can operate stably for a long time to avoid overheating damage. By monitoring these indicators in real time, abnormal operation of the plasma generator can be discovered in time, providing a basis for subsequent predictions and decisions.

[0073] Environmental conditions also have an important impact on the performance of plasma generators, including ambient temperature, humidity, air pressure, and electromagnetic interference. Ambient temperature affects the stability and efficiency of plasma generation. Excessive humidity may cause electrode corrosion or increase in gas humidity, affecting plasma quality. Changes in air pressure will affect the degree of gas ionization and plasma density. Electromagnetic interference may interfere with the control system of the plasma generator and affect its normal operation. Including these environmental conditions in the monitoring range can more accurately predict the operating status of the plasma generator and take corresponding measures to deal with environmental changes.

[0074] The configuration parameters of a plasma generator determine its operating characteristics and performance range. The configuration parameters include the type, size, operating frequency and electrode material of the plasma generator. The type determines how the plasma is generated and the application field. The size affects the volume and shape of the plasma, which in turn affects its application effect. The operating frequency determines the generation rate and energy distribution of the plasma. The electrode material is related to the durability of the electrode and the quality of the plasma. By real-time monitoring and adjusting these configuration parameters, the performance of the plasma generator can be optimized to meet different application requirements.

[0075] In order to achieve real-time monitoring and predictive maintenance of the plasma generator, the present invention defines a series of executable monitoring or adjustment operation sets, including but not limited to:

[0076] Adjusting input power or operating frequency: By changing the input power or operating frequency, the plasma generation rate and energy distribution can be adjusted, thereby optimizing the performance of the plasma generator.

[0077] Changing gas type or flow rate: Depending on application requirements, different gas types and flow rates can be selected to obtain the desired plasma characteristics.

[0078] Start or stop the cooling system: According to the temperature of the plasma generator, start or stop the cooling system in time to ensure its long-term stable operation.

[0079] Adjust the electrode spacing or replace the electrodes: The electrode spacing and the choice of electrode material will affect the quality and stability of the plasma. By adjusting the electrode spacing or replacing the electrodes, the performance of the plasma generator can be optimized.

[0080] Cleaning or replacing parts: Regularly cleaning and replacing parts of the plasma generator can keep it in good working condition and extend its service life.

[0081] Adjust control system parameters: According to actual needs, adjust the control system parameters to achieve more precise plasma generator control.

[0082] Adjust data collection frequency or monitoring accuracy: According to the application scenario and monitoring needs, the frequency of data collection and the monitoring accuracy can be adjusted to meet different monitoring requirements.

[0083] Embodiment 2:

[0084] This embodiment is used to describe in detail the composition of the reward function and the training process of the state monitoring decision model in the real-time monitoring method of the plasma generator, which specifically includes:

[0085] Define the reward function: The reward function R is the core part of the state monitoring decision model. It consists of multiple sub-reward items, and the total reward value is obtained by weighted summation. The specific formula is:

[0086] R=w1*R_stab+w2*R_perf+w3*R_eff+w4*R_maint+w5*R_dev Where: R_stab represents the reward item for state stability, which is calculated according to the degree of fluctuation of the operating state of the plasma generator. If the operating state is stable and the fluctuation is small, a positive reward is given; conversely, if the fluctuation is large, a negative reward or zero reward is given. R_perf represents the reward item for performance maintenance, which is calculated according to the maintenance of the key performance indicators of the plasma generator. If the performance indicator remains within the set range, a positive reward is given; if the performance indicator deviates from the set range, a negative reward is given. R_eff represents the reward item for energy efficiency improvement, which is calculated according to the degree of improvement in energy consumption efficiency. If the energy efficiency is improved, a positive reward is given; if the energy efficiency is reduced, a negative reward is given. R_maint represents the reward item for preventive maintenance, which is calculated according to the execution and effect of preventive maintenance operations. If effective preventive maintenance operations are performed, such as replacing aging parts in advance, a positive reward is given. R_dev represents the reward item or penalty item for performance deviation reduction, which is calculated based on the monitored performance deviation data. Positive rewards are given to operations that reduce the deviation, and the reward value is proportional to the amount of deviation reduction; penalties are given to operations that increase the deviation, and the penalty value is proportional to the amount of deviation increase.

[0087] The weight coefficients w1, w2, w3, w4 and w5 are used to balance the importance of each sub-reward item and satisfy w1+w2+w3+w4+w5=1. These weight coefficients can be adjusted according to the actual application scenario and monitoring requirements.

[0088] In order to train the state monitoring decision model, the present invention adopts the deep Q network (DQN) algorithm. The training process of the state monitoring decision model includes the following steps:

[0089] A1. Initialize the Q network: Build a deep neural network to evaluate the value of state-action pairs. The input of the network is the current state of the plasma generator (including operating indicators, environmental conditions, and configuration parameters), and the output is the value estimate of each possible action.

[0090] A2. Initialize the experience playback buffer D: Set the buffer size N to store the state transition experience generated during the plasma generator monitoring process. Each experience includes the current state s, the selected action a, the reward r, the next state s' and the termination flag done.

[0091] A3. For each training round:

[0092] A301: Initialize the state s at the beginning of the round.

[0093] A302: Before the round ends, the following steps are executed in a loop:

[0094] i) The Q network selects action a according to the current state s, using the ε-greedy strategy to balance exploration and exploitation. That is, it randomly selects actions with a probability of ε, and selects the action with the highest value with a probability of 1-ε.

[0095] ii) Execute action a, observe the reward r and next state s', and whether the termination flag is done.

[0096] iii) Store the experience (s, a, r, s', done) into the experience replay buffer D.

[0097] iv) Randomly sample a batch of experiences from the experience replay buffer D to update the network parameters. This can break the temporal correlation between experiences and improve training efficiency.

[0098] v) Using the sampled experience, update the parameters of the Q network to minimize the value function prediction error. Usually the mean square error or cross entropy loss function is used for training.

[0099] vi) Update the current state s to the next state s'.

[0100] A303: When the round ends (such as reaching the maximum number of steps or the plasma generator fails), reset the state s to the new round start state.

[0101] A4. Regular evaluation and adjustment: During the training process, the performance of the state monitoring decision model is regularly evaluated. Based on the evaluation results, the weight coefficient of the reward function, the learning rate of the model, and the size of the exploration noise are adjusted to optimize the model performance.

[0102] A5. Repeat training: Repeat steps A3 and A4 until a satisfactory performance level is achieved. This usually means that the model is able to stably select the optimal action, keep the plasma generator in an efficient and stable operating state, and reduce performance deviations and failures.

[0103] The DQN algorithm combines the advantages of deep learning and reinforcement learning, and can handle problems in high-dimensional state space and continuous action space.

[0104] Embodiment 3:

[0105] This embodiment is used to describe in detail the specific implementation steps of using the mean square error MSE as the performance evaluation indicator in step A4:

[0106] Step A4: Performance evaluation and parameter update

[0107] In the process of training the state monitoring decision model, in order to evaluate the performance of the current round of training and optimize the model parameters, the mean square error (MSE) is used as the evaluation indicator. The specific implementation steps are as follows:

[0108] A401: Calculate the target Q value: Based on the sampled experience tuple (s, a, r, s'), use the target Q network to calculate the target Q value of each experience tuple.

[0109] A402: Calculate predicted Q-value: Calculate the predicted Q-value of the main Q network for each state-action pair (s, a) in the experience tuple.

[0110] A403: Calculate mean square error (MSE): Calculate the mean square error (MSE) between the predicted Q value and the target Q value as the loss function value of the current iteration. The calculation formula of MSE is:

[0111]

[0112] Among them, N is the number of sampled experience tuples, Q target is the target Q value, Q predicted The MSE value reflects the deviation between the Q value predicted by the main Q network and the target Q value, and is an important indicator for evaluating model performance.

[0113] A404: Parameter update: Based on the calculated MSE value, the parameters of the main Q network are updated using the gradient descent algorithm to minimize the MSE value. The gradient descent algorithm calculates the gradient of the loss function with respect to the network parameters and updates the parameters in the opposite direction of the gradient, thereby gradually reducing the prediction error and optimizing the model performance.

[0114] During the parameter update process, you can set the learning rate α to control the step size of the parameter update. The choice of learning rate has an important impact on the training effect and convergence speed of the model. Too large a learning rate may cause the model to oscillate during the training process, while too small a learning rate may cause the model to converge too slowly or even fail to converge.

[0115] A405: Repeat training: Repeat steps A401 to A404 until the end condition of the current training round is met. The end condition can be reaching a preset number of training rounds, the MSE value is lower than a certain threshold, or the model performance is no longer significantly improved. When the end condition is met, enter the next training round to continue optimizing the model performance.

[0116] Embodiment 4:

[0117] Regarding the construction and training of the operation index prediction model, this embodiment is used to describe the specific implementation steps and structural details of using the gated recurrent unit (GRU) algorithm to construct the prediction model, specifically including:

[0118] ① The operation index prediction model is constructed using the GRU algorithm, and its network structure includes input layer, hidden layer and output layer:

[0119] Input layer: used to receive feature data sets as input. Feature data sets include time series performance indicator data (such as temperature, pressure, current, etc.), environmental parameter data (such as ambient temperature, humidity, etc.), system configuration data (such as device model, setting parameters, etc.) and time-related features (such as timestamp, date, etc.). After preprocessing, these data are input into the GRU network in sequence form.

[0120] Hidden layer: contains at least one GRU hidden layer. Each GRU hidden layer consists of several GRU units, each of which has an update gate and a reset gate. The update gate controls the influence of the previous state on the current state, and the reset gate controls the reset of the previous state on the current input information. These gating mechanisms enable the GRU network to capture long-term dependencies in time series data and effectively handle noise and irregularities in the sequence.

[0121] Output layer: Receives the output of the hidden layer and maps the output of the GRU network to the predicted performance indicator value through a fully connected layer or regression layer. The output layer can be designed as a single output or multiple outputs according to specific needs to predict the future value of one or more performance indicators.

[0122] ② The steps of training the operation indicator prediction model include the following stages:

[0123] S1: Initialize network parameters

[0124] Initialize the weights and biases of the GRU network, usually with small random numbers, to ensure that the model has sufficient exploratory power in the early stages of training.

[0125] S2: Choose a loss function

[0126] Select an appropriate loss function to measure the difference between the model's predicted value and the true value. For regression problems, commonly used loss functions include mean square error (MSE), root mean square error (RMSE), or mean absolute error (MAE). Select an appropriate loss function based on actual needs and data characteristics.

[0127] S3: Iteratively train the network

[0128] The back propagation algorithm and gradient descent optimization algorithm are used for model training. The specific steps are as follows:

[0129] Forward propagation: Input the feature dataset into the GRU network and calculate the network's predicted output.

[0130] Calculate the loss: Calculate the difference between the predicted value and the true value according to the loss function to get the loss value.

[0131] Back propagation: Calculate the gradient based on the loss value, that is, the partial derivative of the loss function with respect to the network parameters.

[0132] Parameter update: Use the gradient descent optimization algorithm Adam to update the weights and bias items of the GRU network to reduce the loss value.

[0133] Repeat the above steps and iteratively train the network until the loss function converges or the set number of training rounds is reached.

[0134] S4: Save the model

[0135] The trained GRU network model is saved as the operation index prediction model in the real-time monitoring method of the plasma generator. The model can be deployed in the monitoring system to receive feature data in real time and predict the future value of the performance index, providing a basis for monitoring decisions.

[0136] The present invention also includes a real-time monitoring system for a plasma generator, the system comprising:

[0137] A data collection module is used to collect the time series data of the operating indicators of the plasma generator in the historical operation, the environmental conditions during the operation and the configuration parameters of the plasma generator, and pre-process the collected data to form a feature data set;

[0138] A prediction model building module is used to build an operation index prediction model based on a feature data set, combined with environmental conditions and plasma generator configuration parameters, so as to predict the change trend of the operation index of the plasma generator in the future;

[0139] The prediction data acquisition module is used to obtain the current configuration parameter data and environmental condition data of the plasma generator in real time, and input the acquired data into the operation index prediction module, and receive and output the operation index predicted by the model;

[0140] A real-time data acquisition module is used to obtain real-time operating index data of the plasma generator;

[0141] Deviation detection module, used to draw the real-time operation index data and the predicted operation index data into time-varying curves, align the timestamps and place them in the same coordinate system for comparison, further set the performance deviation threshold, determine whether each corresponding point on the two curves has a deviation exceeding the threshold, and mark the deviation point;

[0142] The state monitoring decision module is used to build a state monitoring decision model, modeling the state monitoring process of the plasma generator as a Markov decision process, including defining the state space, action space, reward function and output strategy, and continuously optimizing the generated adjustment actions through the reinforcement learning algorithm;

[0143] The decision execution module is used to perform real-time monitoring based on the trained state monitoring decision model, collect the configuration parameters and environmental condition data of the plasma generator in real time as model input, select the optimal adjustment action according to the current state and strategy, and automatically execute the decision result or transmit it to the operator for confirmation through the monitoring system.

[0144] The implementation of the system refers to the above embodiment and will not be described in detail in the specification.

[0145] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0146] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time monitoring method for a plasma generator, characterized in that: The method comprises: Collecting the time series data of the operating indicators of the plasma generator in historical operation, as well as the environmental conditions during operation and the configuration parameters of the plasma generator; and preprocessing the collected data to form a feature data set; A prediction model for the operation index of the plasma generator is constructed. The model is based on the characteristic data set, combined with environmental conditions and plasma generator configuration parameters, to predict the change trend of the operation index of the plasma generator in the future. Acquire the current configuration parameter data and environmental condition data of the plasma generator in real time, input the acquired data into the operation index prediction model, and receive the predicted operation index output by the model; Acquire real-time operating index data of the plasma generator; plot the acquired real-time operating index data and the predicted operating index data output by the model into time-varying curves, align the timestamps of the two curves, and place the two curves in the same coordinate system for comparison; further set a performance deviation threshold to determine whether each corresponding point on the two curves has a deviation exceeding the threshold, and if so, mark it as a deviation point; A state monitoring decision model is constructed, and the state monitoring process of the plasma generator is modeled as a Markov decision process. The specific definitions include: ① State space: defined as the combination of real-time operating indicator data, configuration parameters and environmental condition data of the plasma generator; ② Action space: defined as a set of executable monitoring or adjustment operations; ③ Reward function: designed according to the stable operation and performance optimization goals of the plasma generator, including reward mechanisms for state stability, performance maintenance and energy efficiency improvement. At the same time, combined with the monitored performance deviation data, additional rewards are given to operations that reduce deviations, and additional penalties are given to operations that increase deviations; ④ Output strategy: defined as a method for selecting the optimal action under the current state, and the generated adjustment actions are continuously optimized through reinforcement learning algorithms; Real-time monitoring is performed based on the trained state monitoring decision model. The configuration parameters and environmental condition data of the plasma generator are collected in real time as model input. The optimal adjustment action is selected according to the current state and strategy, and the decision result is automatically executed or transmitted to the operator through the monitoring system for confirmation before execution.

2. The real-time monitoring method of a plasma generator according to claim 1, characterized in that: The operating indicators include plasma density, temperature, voltage, current, power, gas flow and cooling system status; the environmental conditions include ambient temperature, humidity, air pressure and electromagnetic interference; the configuration parameters include the type, size, operating frequency and electrode material of the plasma generator.

3. The real-time monitoring method of a plasma generator according to claim 1, characterized in that: The executable monitoring or adjustment operation set includes: adjusting input power or operating frequency, changing gas type or flow, starting or stopping cooling system, adjusting electrode spacing or replacing electrodes, cleaning or replacing parts, adjusting control system parameters, adjusting data acquisition frequency or monitoring accuracy.

4. The real-time monitoring method of a plasma generator according to claim 1, characterized in that: The reward function R is composed of multiple sub-reward items, and the total reward value is obtained by weighted summation. The specific formula is: R = w1*R_stab+w2*R_perf+w3*R_eff+w4*R_maint+w5*R_dev; where R_stab represents the reward item for state stability, which is calculated according to the fluctuation degree of the operating state of the plasma generator; R_perf represents the reward item for performance maintenance, which is calculated according to the maintenance of the key performance indicators of the plasma generator; R_eff represents the reward item for energy efficiency improvement. Item, calculated according to the degree of improvement in energy consumption efficiency; R_maint represents the reward item for preventive maintenance, calculated according to the execution and effect of preventive maintenance operations; R_dev represents the reward item or penalty item for reducing performance deviation, and according to the monitored performance deviation data, positive rewards are given to operations that reduce deviations, and penalties are given to operations that increase deviations, and its value is set as a function that is inversely proportional or directly proportional to the amount of deviation change; w1, w2, w3, w4 and w5 are the weight coefficients of each sub-reward item, and w1+w2+w3+w4+w5=1.

5. The real-time monitoring method of a plasma generator according to claim 4, characterized in that: The operation index prediction model is constructed using the gated recurrent unit GRU algorithm, and its network structure specifically includes: Input layer: used to receive feature data sets as input, wherein the feature data sets include time series performance indicator data, environmental parameter data, system configuration data, and time-related features; Hidden layer: contains at least one GRU hidden layer. Each GRU hidden layer contains several GRU units. Each GRU unit has an update gate and a reset gate to control the flow of information and state updates. Output layer: Receives the output of the hidden layer and maps the output of the GRU network to the predicted performance indicator value through a fully connected layer or regression layer.

6. The real-time monitoring method of a plasma generator according to claim 5, characterized in that: The steps to train the running indicator prediction model include: S1: Initialize the weights and biases of the GRU network using small random numbers; S2: Select the loss function to measure the difference between the model prediction value and the true value; S3: Use the back propagation algorithm and gradient descent optimization algorithm to calculate the gradient according to the loss function, update the weights and bias items of the GRU network, and iteratively train the network until the loss function converges; S4: Save the trained GRU network model as an operation indicator prediction model in the real-time monitoring method of the plasma generator.

7. The real-time monitoring method of a plasma generator according to claim 4, characterized in that: The deep Q network DQN algorithm is used to train the state monitoring decision model.

8. The real-time monitoring method of a plasma generator according to claim 7, characterized in that: The steps to train a condition monitoring decision model include: A1. Initialize the Q network: build a deep neural network to evaluate the value of state-action pairs; A2. Initialize the experience playback buffer D: Set the buffer size N to store the state transition experience generated during the plasma generator monitoring process, including state s, action a, reward r, next state s' and whether the termination flag is done; A3. For each training round: A301: Initialize the state s at the beginning of the round; A302: Before the round ends, the following steps are executed in a loop: i) The Q network selects action a according to the current state s, and adopts the ε-greedy strategy to balance exploration and utilization; ii) Execute action a, observe the reward r and the next state s', and whether the termination flag is done; iii) Store the experience (s, a, r, s', done) in the experience replay buffer D; iv) Randomly sample a batch of experience from the experience replay buffer D to update the network parameters; v) Use the sampled experience to update the parameters of the Q network to minimize the prediction error of the value function; vi) Update the current state s to the next state s'; A303: When the round ends, reset state s to the new round start state; A4: During the training process, the performance of the state monitoring decision model is regularly evaluated, and the weight coefficient of the reward function, the learning rate of the model, and the size of the exploration noise are adjusted based on the evaluation results; A5: Repeat steps A3 and A4 until a satisfactory performance level is achieved.

9. The real-time monitoring method of a plasma generator according to claim 8, characterized in that: In step A4, mean square error (MSE) is used as the performance evaluation index of the current round of training, and the specific implementation steps include: A401: Based on the sampled experience tuple (s, a, r, s'), the target Q value of each experience tuple is calculated using the target Q network; A402: Calculate the predicted Q value of the main Q network for each state-action pair (s, a) in the experience tuple; A403: Calculate the mean square error (MSE) between the predicted Q value and the target Q value as the loss function value of the current iteration. The calculation formula of MSE is: Among them, N is the number of sampled experience tuples, Q target is the target Q value, Q predicted is the predicted Q value of the main Q network; A404: Based on the calculated MSE value, the parameters of the main Q network are updated using the gradient descent algorithm to minimize the MSE value, thereby optimizing the performance of the main Q network; A405: Repeat steps A401 to A404 until the end condition of the current training round is met, and then enter the next training round.

10. A real-time monitoring system for a plasma generator, characterized in that: The system comprises: A data collection module is used to collect the time series data of the operating indicators of the plasma generator in the historical operation, the environmental conditions during the operation and the configuration parameters of the plasma generator, and pre-process the collected data to form a feature data set; A prediction model building module is used to build an operation index prediction model based on a feature data set, combined with environmental conditions and plasma generator configuration parameters, so as to predict the change trend of the operation index of the plasma generator in the future; The prediction data acquisition module is used to obtain the current configuration parameter data and environmental condition data of the plasma generator in real time, and input the acquired data into the operation index prediction module, and receive and output the operation index predicted by the model; A real-time data acquisition module is used to obtain real-time operating index data of the plasma generator; Deviation detection module, used to draw the real-time operation index data and the predicted operation index data into time-varying curves, align the timestamps and place them in the same coordinate system for comparison, further set the performance deviation threshold, determine whether each corresponding point on the two curves has a deviation exceeding the threshold, and mark the deviation point; The state monitoring decision module is used to build a state monitoring decision model, modeling the state monitoring process of the plasma generator as a Markov decision process, including defining the state space, action space, reward function and output strategy, and continuously optimizing the generated adjustment actions through the reinforcement learning algorithm; The decision execution module is used to perform real-time monitoring based on the trained state monitoring decision model, collect the configuration parameters and environmental condition data of the plasma generator in real time as model input, select the optimal adjustment action according to the current state and strategy, and automatically execute the decision result or transmit it to the operator for confirmation through the monitoring system.

Images