Intelligent operation and maintenance method and system for photo-thermal power generation system based on Internet of Things

Abstract

The invention provides an intelligent operation and maintenance method and system for a photo-thermal power generation system based on the Internet of Things, and relates to the technical field of information technologies, and the method comprises the steps: obtaining temperature, humidity and vibration environment parameter data collected by a sensor disposed on power generation equipment; and receiving the environmental parameter data, comparing the environmental parameter data with a preset threshold value, and judging whether the environmental parameters exceed a normal range or not. According to the method, environmental parameters such as temperature, humidity and vibration of the equipment are collected in real time by arranging the sensors, when the parameters exceed the preset threshold value, performance parameter collection is triggered, weighted average calculation is carried out in combination with the environmental parameters, and the equipment health index is obtained; according to the method, intelligent assessment and predictive maintenance of the health state of the power generation equipment are realized, the operation reliability and the maintenance efficiency of the equipment are improved, the equipment fault risk is reduced, and the service life of the equipment is prolonged.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an intelligent operation and maintenance method and system for a photothermal power generation system based on the Internet of Things. Background Art

[0002] In the operation and maintenance of power generation equipment, real-time monitoring of environmental conditions is crucial. When parameters such as temperature, humidity, and vibration exceed preset thresholds, early warning information needs to be generated in a timely manner to prompt operation and maintenance personnel to take necessary protective measures to ensure the safe operation of the equipment. However, frequent early warning information may cause alarm fatigue for operation and maintenance personnel, resulting in important warnings being ignored. At the same time, too frequent equipment maintenance will also cause unnecessary interference to the power generation process and reduce power generation efficiency.

[0003] How to minimize the interference of early warning information on operation and maintenance personnel and optimize maintenance scheduling while ensuring the safe operation of equipment has become a technical problem that needs to be solved urgently. Traditional rule-based early warning methods are difficult to adapt to complex and changing environmental conditions, and simple time interval-based maintenance scheduling cannot fully consider the actual health status of the equipment. This requires the introduction of a more intelligent and personalized early warning generation and maintenance scheduling mechanism. Through in-depth analysis of equipment historical data and environmental conditions, a dynamic model of equipment health status is established to achieve accurate generation of early warning information and dynamic optimization of maintenance time windows. While ensuring equipment safety, the impact on the power generation process is minimized, and power generation efficiency and economic benefits are improved. Summary of the invention

[0004] The present invention provides an intelligent operation and maintenance method for a solar thermal power generation system based on the Internet of Things, which mainly includes:

[0005] Obtain temperature, humidity, and vibration environmental parameter data collected by sensors deployed on power generation equipment;

[0006] Receive environmental parameter data, compare the environmental parameter data with preset thresholds, and determine whether the environmental parameters exceed a normal range;

[0007] When environmental parameters exceed preset thresholds, an early warning message containing the abnormal parameter type, value, and occurrence time is generated, and the priority is evaluated based on the early warning message;

[0008] According to the priority of the early warning information, the power output, fuel consumption rate, and vibration frequency performance parameters of the power generation equipment are collected, and the health status of the power generation equipment is evaluated by combining the environmental parameter data and performance parameters using a weighted average algorithm to obtain a health index.

[0009] When the health index of the power generation equipment is lower than the preset threshold, a maintenance plan is generated, which includes maintenance items, required personnel, tools, spare parts and estimated time;

[0010] Track the progress of maintenance plan execution, obtain equipment performance parameters after maintenance, and use machine learning algorithms to update the health status model of power generation equipment based on performance parameters and environmental parameters, including adjusting the weight coefficients of each parameter;

[0011] Based on the updated health status model, a dynamic threshold adjustment algorithm is used to automatically adjust the preset thresholds of environmental parameters and performance parameters to optimize the early warning mechanism and maintenance strategy.

[0012] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things of the present invention, the temperature, humidity and vibration environmental parameter data collected by the sensors deployed on the power generation equipment are obtained, including:

[0013] Obtaining environmental parameter data collected by temperature, humidity and vibration sensors on the power generation equipment, and determining whether the data exceeds a preset threshold;

[0014] If exceeded, the current operation data and historical operation data of the power generation equipment are obtained, an isolation forest algorithm is used to perform anomaly detection on the current operation data, and the detection result is compared with the historical operation data to determine the specific cause of the anomaly;

[0015] According to the abnormal cause, relevant operating parameters are extracted as decision tree model features, and the decision tree model is trained to generate corresponding operation and maintenance strategies.

[0016] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things of the present invention, wherein: the receiving of environmental parameter data, comparing the environmental parameter data with a preset threshold value, and judging whether the environmental parameter exceeds a normal range includes:

[0017] Obtain the preprocessed environmental parameter data and preset thresholds, and use the cosine similarity algorithm to calculate the similarity score;

[0018] If the similarity score is lower than the threshold, an abnormal alarm is triggered;

[0019] Obtain the severity and duration of abnormal situations and generate emergency response plans through the CART decision tree algorithm;

[0020] The exception handling process and result data are stored in the database, and the historical data is analyzed using the LSTM neural network algorithm to dynamically adjust the threshold settings and optimize the monitoring system.

[0021] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things of the present invention, when the environmental parameters exceed the preset threshold, a warning message including the abnormal parameter type, value and occurrence time is generated, and the priority is evaluated according to the warning information, including:

[0022] Acquire various environmental parameter data collected by the sensor in real time, and send the environmental parameter data to the data processing module through the wireless transmission module;

[0023] Determine whether the environmental parameter data exceeds a preset threshold value, and if so, trigger a pre-established abnormality determination mechanism;

[0024] According to the priority level of the warning information and in combination with a preset push strategy, a push method and a push channel for the warning information are determined, and the warning information is pushed to at least one target terminal.

[0025] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things of the present invention, the priority of the early warning information is used to trigger the collection of the power output, fuel consumption rate, and vibration frequency performance parameters of the power generation equipment, and the health status of the power generation equipment is evaluated by combining the environmental parameter data and the performance parameters using a weighted average algorithm to obtain a health index, including:

[0026] Obtaining the weight coefficients of the preset performance parameters and environmental parameters;

[0027] The weighted average algorithm is used to perform weighted calculation on the collected performance parameters and environmental parameters to obtain comprehensive evaluation indicators;

[0028] Comparing the comprehensive evaluation index with a preset health status threshold to determine the health status level of the power generation equipment;

[0029] According to the health status level, combined with a preset health index calculation formula, calculate the health index value of the power generation equipment;

[0030] If the health index value is higher than a preset threshold, it is determined that the power generation equipment is in good operating condition;

[0031] Otherwise, it is determined that there is a potential failure risk.

[0032] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things described in the present invention, when the health index of the power generation equipment is lower than the preset threshold, a maintenance plan is generated, and the maintenance plan includes maintenance items, required personnel, tools, spare parts and estimated time, including:

[0033] Acquire real-time health index data of the power generation equipment and determine whether the health index is lower than a preset threshold;

[0034] If the health index is lower than the threshold, a CART decision tree algorithm is used to generate a maintenance item list by taking equipment parameters as features and whether maintenance is required as the target variable and training based on historical maintenance data;

[0035] According to the maintenance project list, obtain required resource information;

[0036] By connecting with the ERP system, real-time inventory of resources can be obtained. If resources are insufficient, the priority of the maintenance plan can be adjusted according to the urgency of the maintenance project to optimize the feasibility of the maintenance plan.

[0037] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things of the present invention, wherein: the tracking of the execution progress of the maintenance plan, obtaining the performance parameters of the equipment after maintenance, and updating the health status model of the power generation equipment using a machine learning algorithm according to the performance parameters and environmental parameters, including adjusting the weight coefficient of each parameter, include:

[0038] Obtain equipment performance data and environmental parameter data and input them into a pre-built equipment health status assessment model trained using a support vector machine algorithm;

[0039] According to the difference between the health status value calculated by the model and the actual equipment operation status, a grid search method is used to adjust the penalty coefficient and kernel function type hyperparameters of the support vector machine model to optimize the model;

[0040] The optimized equipment health status assessment model is deployed to the production environment, and the real-time collected equipment performance data and environmental parameter data are predicted to output the equipment health status prediction results.

[0041] As a preferred solution of the intelligent operation and maintenance method and system of the solar thermal power generation system based on the Internet of Things of the present invention, wherein: according to the updated health status model, the preset thresholds of environmental parameters and performance parameters are automatically adjusted using a dynamic threshold adjustment algorithm to optimize the early warning mechanism and maintenance strategy, including:

[0042] Obtain historical equipment operation data, determine the relationship between various performance parameters and equipment health status through correlation analysis, and establish an equipment health status assessment model;

[0043] For the newly collected real-time operation data of the equipment, the support vector machine algorithm is used for incremental learning, which is added to the original historical operation data, so as to continuously update and improve the equipment health status assessment model. By comparing the equipment performance indicators and maintenance costs before and after optimization, the maintenance strategy is continuously improved.

[0044] An intelligent operation and maintenance system for a photothermal power generation system based on the Internet of Things, based on the intelligent operation and maintenance method for a photothermal power generation system based on the Internet of Things, characterized by comprising an environmental parameter acquisition module, an environmental parameter monitoring and alarm module, an equipment health status assessment module, a maintenance plan generation and optimization module, a model update and optimization module, an anomaly detection and processing module and a model update module;

[0045] Environmental parameter acquisition module, used to obtain temperature, humidity, and vibration environmental parameter data collected by sensors deployed on power generation equipment;

[0046] The environmental parameter monitoring and alarm module is used to receive environmental parameter data, compare the environmental parameter data with the preset threshold value, and determine whether the environmental parameter exceeds the normal range;

[0047] The equipment health status assessment module is used to generate warning information containing abnormal parameter type, value, and occurrence time when environmental parameters exceed the preset threshold, and evaluate the priority according to the warning information;

[0048] The maintenance plan generation and optimization module is used to trigger the collection of power output, fuel consumption rate, and vibration frequency performance parameters of power generation equipment according to the priority of early warning information, and use the weighted average algorithm to combine environmental parameter data and performance parameters to evaluate the health status of the power generation equipment and obtain the health index;

[0049] Model update and optimization module, used to generate maintenance plans when the health index of power generation equipment falls below a preset threshold;

[0050] Anomaly detection and processing module, used to track the progress of maintenance plan execution and obtain equipment performance parameters after maintenance;

[0051] The model update module uses a dynamic threshold adjustment algorithm to automatically adjust the preset thresholds of environmental parameters and performance parameters according to the updated health status model, and optimizes the early warning mechanism and maintenance strategy.

[0052] A computer device comprises: a memory and a processor; the memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the intelligent operation and maintenance method of a solar thermal power generation system based on the Internet of Things are implemented.

[0053] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0054] The present invention discloses a method and system for intelligently evaluating the health status of power generation equipment. Sensors are deployed to collect environmental parameters such as temperature, humidity, and vibration of the equipment in real time. When the parameters exceed the preset threshold, performance parameter collection is triggered and weighted average calculation is performed in combination with the environmental parameters to obtain the equipment health index. When the health index is lower than the threshold, a maintenance plan containing maintenance items, personnel, tools, and the like is automatically generated. After the maintenance is completed, the maintenance effect is analyzed based on a machine learning algorithm, the parameter weights of the health status model are updated, and a dynamic threshold adjustment algorithm is used to optimize the early warning mechanism. The method realizes intelligent evaluation and predictive maintenance of the health status of power generation equipment, improves the equipment operation reliability and maintenance efficiency, reduces the risk of equipment failure, and extends the service life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 The present invention is a flow chart of an intelligent operation and maintenance method and system for a solar thermal power generation system based on the Internet of Things.

[0056] Figure 2 It is a schematic diagram of an intelligent operation and maintenance method and system of a solar thermal power generation system based on the Internet of Things of the present invention.

[0057] Figure 3 This is another schematic diagram of an intelligent operation and maintenance method and system for a solar thermal power generation system based on the Internet of Things of the present invention. DETAILED DESCRIPTION

[0058] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than 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] Example 1

[0060] like Figure 1-Figure 3 In this embodiment, an intelligent operation and maintenance method of a solar thermal power generation system based on the Internet of Things may specifically include:

[0061] S101. Acquire temperature, humidity, and vibration environmental parameter data collected by sensors deployed on power generation equipment.

[0062] Acquire environmental parameter data collected by temperature, humidity and vibration sensors on the power generation equipment, and determine whether the data exceeds a preset threshold; if so, acquire current operating data and historical operating data of the power generation equipment, use an isolation forest algorithm to perform anomaly detection on the current operating data, compare the detection result with the historical operating data, and determine the specific cause of the anomaly; based on the cause of the anomaly, extract relevant operating parameters as features of a decision tree model, and train the decision tree model to generate corresponding operation and maintenance strategies.

[0063] Specifically, the environmental parameter data collected by the temperature, humidity and vibration sensors deployed on the power generation equipment are obtained. According to the preset temperature, humidity and vibration thresholds, it is judged whether the collected data exceeds the normal range. If any parameter exceeds the threshold, it is preliminarily determined that the operating state of the power generation equipment is abnormal. For the power generation equipment preliminarily determined to be abnormal, its operating data in the recent period is obtained as the current operating data; at the same time, historical operating data similar to the current season, time period and other conditions are obtained. Anomaly detection algorithms, such as the isolation forest algorithm, are used to detect anomalies in the current operating data. The detection results are compared with the historical operating data to identify abnormal features that are significantly different from the historical data, so as to determine the specific abnormal causes, such as abnormal temperature rise, abnormal increase in vibration amplitude, etc. Based on the determined abnormal causes, relevant operating parameters, such as temperature, speed, etc., are extracted as features of the decision tree model. The historical operation and maintenance records are used as training data to train the decision tree model, and the operation and maintenance strategies for the abnormal causes are automatically generated, such as adjusting equipment parameters, replacing parts, and shutting down for maintenance. The generated operation and maintenance strategies are sent to relevant operation and maintenance personnel to guide them to repair and debug the abnormal equipment according to the strategies to ensure that the power generation equipment resumes normal operation as soon as possible. After the operation and maintenance is completed, the abnormal situation and the handling process will be recorded and added to the historical operation and maintenance records to optimize the decision tree model and improve the accuracy and effectiveness of future operation and maintenance strategies.

[0064] S102: Receive environmental parameter data, compare the environmental parameter data with a preset threshold, and determine whether the environmental parameter exceeds a normal range.

[0065] Obtain pre-processed environmental parameter data and preset thresholds, and use the cosine similarity algorithm to calculate the similarity score. If the similarity score is lower than the threshold, an abnormal alarm is triggered. Obtain the severity and duration of the abnormal situation, and generate an emergency response plan through the CART decision tree algorithm. Store the abnormal processing process and result data in the database, use the LSTM neural network algorithm to analyze historical data, dynamically adjust the threshold setting, and optimize the monitoring system.

[0066] Specifically, the environmental parameter data collected by multiple sensors are obtained, and different types of environmental parameter data are preprocessed to unify the data format. The preprocessing process includes operations such as removing outliers, smoothing noise, and normalization. The NumPy and Pandas libraries of Python are used for data cleaning and conversion. According to the type of environmental parameter, the normal range threshold of the corresponding type of environmental parameter is obtained from the preset threshold database. The cosine similarity algorithm is used to compare the preprocessed environmental parameter data with the corresponding threshold and calculate the similarity score. The cosine similarity algorithm measures the similarity between two vectors by calculating the cosine value of the angle between them, and the value range is [0,1]. The similarity score threshold is set to determine whether the environmental parameter exceeds the normal range. If the similarity score is lower than the threshold, an abnormal situation alarm is triggered. The alarm is presented on the monitoring system interface in the form of a pop-up window and sound prompt, and text messages and email notifications are sent to relevant personnel. The environmental parameter data and alarm information of the abnormal situation are sent to the monitoring system, and the abnormal data and alarm information are displayed in real time through the visual interface. Matplotlib and Seaborn libraries are used to draw line graphs, scatter plots, etc. to intuitively display the data change trend and the distribution of abnormal points. According to the severity and duration of the abnormal situation, the CART decision tree algorithm is used to automatically generate emergency treatment plans. The severity is divided according to the gap between the similarity score and the threshold, and the duration is divided according to the number of consecutive abnormal states. The severity and duration are used as input features of the decision tree, and the treatment plan is used as the output target to train the decision tree model. When an abnormal situation occurs, its features are input into the decision tree, and the corresponding treatment plan is automatically generated to guide the on-site personnel to deal with it. The processing process and result data of the abnormal situation are stored in the database, and the historical data is analyzed using the LSTM neural network algorithm. The abnormal data and the corresponding processing results are used as training samples to train the LSTM model and learn the correlation between the abnormal situation and the treatment effect. Based on the prediction results of the LSTM model, the threshold setting is dynamically adjusted to optimize the sensitivity and accuracy of the monitoring system. Through the above steps, the full process management of real-time monitoring of environmental parameters, anomaly detection, alarm triggering, emergency treatment and dynamic optimization is realized, which improves the intelligence level and response efficiency of environmental safety monitoring.

[0067] S103. When an environmental parameter exceeds a preset threshold, generate warning information including the abnormal parameter type, value, and occurrence time, and evaluate the priority based on the warning information.

[0068] Acquire various environmental parameter data collected by the sensor in real time, and send the environmental parameter data to the data processing module through the wireless transmission module;

[0069] Determine whether the environmental parameter data exceeds a preset threshold value, and if so, trigger a pre-established abnormality determination mechanism;

[0070] According to the priority level of the warning information and in combination with a preset push strategy, a push method and a push channel for the warning information are determined, and the warning information is pushed to at least one target terminal.

[0071] Specifically, various environmental parameter data are collected in real time through sensors, and the collected data are sent to the data processing module through the wireless transmission module. The data processing module receives the environmental parameter data collected by the sensor through the wireless receiving module and compares it with the preset threshold. If the environmental parameter data exceeds the preset threshold, the abnormal judgment mechanism is triggered. The abnormal judgment mechanism determines the type and specific value of the abnormal parameter by comparing the difference between the current data and the historical data, combined with the abnormal judgment rules in the expert knowledge base. After the abnormal judgment mechanism is triggered, the system automatically records the time when the abnormality occurs, and integrates the abnormal parameter type, value and time information to generate early warning information. According to the pre-established priority evaluation decision tree, the priority of the generated early warning information is determined. Multiple judgment nodes are set in the decision tree, and the priority of the early warning information is determined according to factors such as the severity, impact range, and duration of the early warning information. The content and priority of the early warning information are transmitted to the early warning information release module. The early warning information release module automatically determines the push method and channel of the early warning information according to the priority of the early warning information and the preset push strategy. High-priority early warning information will be quickly pushed to relevant personnel through instant messaging tools, text messages, etc.; general priority early warning information will be sent through emails, APP push, etc. At the same time, the system intelligently selects the optimal push channel based on the receiving devices and personal preferences of relevant personnel to ensure the effective communication of early warning information.

[0072] S104. According to the priority of the early warning information, the power output, fuel consumption rate, and vibration frequency performance parameters of the power generation equipment are collected, and a weighted average algorithm is used to combine the environmental parameter data and the performance parameters to evaluate the health status of the power generation equipment and obtain a health index.

[0073] Obtaining the weight coefficients of the preset performance parameters and environmental parameters;

[0074] The weighted average algorithm is used to perform weighted calculation on the collected performance parameters and environmental parameters to obtain comprehensive evaluation indicators;

[0075] Comparing the comprehensive evaluation index with a preset health status threshold to determine the health status level of the power generation equipment;

[0076] According to the health status level, combined with a preset health index calculation formula, calculate the health index value of the power generation equipment;

[0077] If the health index value is higher than a preset threshold, it is determined that the power generation equipment is in good operating condition; otherwise, it is determined that there is a potential failure risk.

[0078] Specifically, if a high-priority early warning message is received, the performance parameters of the power generation equipment are collected, including parameters such as power output, fuel consumption rate and vibration frequency. At the same time, the environmental parameter data such as temperature, humidity, and air pressure of the current power generation equipment are obtained through environmental sensors. According to expert experience, the weight coefficients of various performance parameters and environmental parameters are pre-set. The weighted average algorithm is used to perform weighted calculation on the collected performance parameters and environmental parameters to obtain a comprehensive evaluation index. The formula of the weighted average algorithm is: Comprehensive evaluation index = ∑ (parameter value i × weight coefficient i) The comprehensive evaluation index is compared with the preset health status threshold to determine the health status level of the current power generation equipment. If the comprehensive evaluation index is higher than the preset health status threshold, the health status level of the current power generation equipment is marked as "healthy"; if the comprehensive evaluation index is lower than the preset health status threshold, the health status level of the current power generation equipment is marked as "sub-healthy" or "unhealthy". According to the health status level of the power generation equipment, combined with the preset health index calculation formula, the health index value of the current power generation equipment is calculated. The health index calculation formula is: Health Index = Health Status Level × 100. The calculated health index value is compared with the preset health index threshold. If the health index value is higher than the preset threshold, it is determined that the current power generation equipment is in good operating condition; if the health index value is lower than the preset threshold, it is determined that the current power generation equipment has a potential failure risk and needs maintenance or overhaul. Based on the judgment result of the health index, a health report of the power generation equipment is automatically generated and sent to the relevant operation and maintenance personnel to prompt the current health status and potential risks of the equipment, providing a reference for subsequent operation and maintenance decisions.

[0079] S105. When the health index of the power generation equipment is lower than a preset threshold, a maintenance plan is generated, where the maintenance plan includes maintenance items, required personnel, tools, spare parts, and estimated time.

[0080] Acquire real-time health index data of power generation equipment to determine whether the health index is lower than a preset threshold; if the health index is lower than the threshold, use the CART decision tree algorithm, with equipment parameters as features and whether maintenance is required as the target variable, train based on historical maintenance data, and generate a maintenance project list; obtain required resource information based on the maintenance project list; obtain real-time inventory of resources by connecting with the ERP system; if resources are insufficient, adjust the priority of the maintenance plan according to the urgency of the maintenance project to optimize the executability of the maintenance plan.

[0081] Specifically, the real-time health index data of the power generation equipment is obtained, and the health index threshold specified in the State Grid Enterprise Standard DL / T1092-2012 "Power Equipment Status Maintenance Guidelines" is used to determine whether the equipment health index is lower than the threshold. If the health index is lower than the preset threshold, the CART decision tree algorithm is used based on the historical maintenance data and current status parameters of the power generation equipment, with equipment type, operating time, failure frequency, etc. as features, and whether maintenance is required as the target variable. It is trained based on the historical maintenance record data set within the enterprise to automatically generate a targeted maintenance project list. According to the generated maintenance project list, by querying the structured maintenance knowledge base stored in the MySQL database, the information such as the personnel role, professional tools and spare parts model required for each maintenance project is obtained. According to the complexity of the maintenance project and the historical maintenance working hours data, the multivariate linear regression algorithm is used, with the complexity of the maintenance project, the number of required personnel, the number of required tools, the number of spare parts replacement, etc. as independent variables, and the historical maintenance working hours as the dependent variable, to establish a regression model to estimate the expected completion time of each maintenance project. The generated maintenance project, required personnel, tools, spare parts and expected time information are integrated to form a complete maintenance plan. By connecting with the SAPERP system, we can obtain the real-time inventory and deployment of personnel, tools and spare parts. If resources are insufficient, we can adjust the priority and schedule of the maintenance plan according to the urgency and impact of the maintenance project to optimize the feasibility of the maintenance plan. The optimized maintenance plan will automatically generate work orders through the OA system and push them to the maintenance personnel's DingTalk mobile application to track the execution progress and result feedback of each maintenance project in real time. For projects with problems or delays, we will provide timely warnings and dispatch to ensure that maintenance work is completed efficiently and improve the health level and reliability of power generation equipment.

[0082] S106. Track the execution progress of the maintenance plan, obtain the performance parameters of the equipment after maintenance, and use the machine learning algorithm to update the health status model of the power generation equipment according to the performance parameters and environmental parameters, including adjusting the weight coefficient of each parameter.

[0083] The equipment performance data and environmental parameter data are obtained and input into a pre-built equipment health status assessment model trained with a support vector machine algorithm; in view of the difference between the health status value calculated by the model and the actual equipment operation status, the penalty coefficient and kernel function type hyperparameters of the support vector machine model are adjusted using a grid search method to optimize the model; the optimized equipment health status assessment model is deployed to the production environment, the real-time collected equipment performance data and environmental parameter data are predicted, and the equipment health status prediction result is output.

[0084] Specifically, according to the equipment maintenance plan, the execution progress and completion status of each maintenance task are recorded to obtain the progress data of the maintenance work. After the maintenance is completed, the performance parameters of the equipment, such as temperature, pressure, speed, etc., are collected through sensors to form equipment performance data. At the same time, the temperature, humidity, air pressure and other parameters of the equipment operating environment are obtained as external factors affecting the health status of the equipment. The equipment performance data and environmental parameter data are input into the pre-built equipment health status assessment model, which is trained using the support vector machine (SVM) algorithm. The current health status value of the equipment is obtained through model calculation. According to the difference between the health status value calculated by the model and the actual equipment operating status, the hyperparameters of the SVM model, such as the penalty coefficient, kernel function type, etc., are adjusted using methods such as grid search to optimize the model to make the output results more accurate. If the performance indicators of the optimized model on the validation data set, such as mean square error (MSE), determination coefficient (R 2 ), etc., if the preset target value is not reached, consider using other machine learning algorithms, such as random forest, gradient boosting decision tree (GBDT), etc., to retrain the model, compare the performance of different models, and select the model with the best effect. Deploy the trained equipment health status assessment model to the production environment to predict the equipment performance data and environmental parameter data collected in real time. According to the health status value output by the model, set the early warning threshold. When the health status of the equipment is lower than the threshold, generate an alarm message to notify relevant personnel to perform equipment inspection or maintenance to prevent equipment failure in advance. At the same time, continue to collect real-time health status data during equipment operation to evaluate the actual prediction effect of the model. Regularly use newly collected data to retrain and optimize the model to continuously improve the accuracy and reliability of equipment health status assessment.

[0085] S107. According to the updated health status model, a dynamic threshold adjustment algorithm is used to automatically adjust preset thresholds of environmental parameters and performance parameters to optimize the early warning mechanism and maintenance strategy.

[0086] Obtain the historical operation data of the equipment, determine the relationship between various performance parameters and the health status of the equipment through correlation analysis, and establish an equipment health status assessment model; use the support vector machine algorithm to perform incremental learning on the newly collected real-time operation data of the equipment, add it to the original historical operation data, continuously update and improve the equipment health status assessment model, and continuously improve the maintenance strategy by comparing the equipment performance indicators and maintenance costs before and after optimization.

[0087] A computer device comprises: a memory and a processor; the memory stores a computer program, and when the processor executes the computer program, the steps of the intelligent operation and maintenance method of a solar thermal power generation system based on the Internet of Things are implemented.

[0088] Specifically, according to the historical operation data of the equipment, the relationship between each performance parameter and the health status of the equipment is determined through correlation analysis, and an equipment health status evaluation model is established. The real-time operation data of the equipment is evaluated by using the model to calculate the current health status score of the equipment. The normal range thresholds of each performance parameter are preset. When the real-time operation data of the equipment exceeds the threshold range, the early warning mechanism is triggered. According to the health status score given by the health status evaluation model, the early warning level is judged, and the corresponding equipment maintenance strategy is adopted. The equipment performance is optimized by automatically adjusting the equipment operating environment parameters to restore it to a normal level. The decision tree algorithm is used to analyze the historical operation data of the equipment to obtain the optimal operating parameters of the equipment under different health conditions. These parameters are used as the basis for dynamically adjusting the preset threshold range. During the operation of the equipment, the real-time operation data of the equipment is continuously collected. The support vector machine algorithm is used for incremental learning, and the newly collected data is added to the original historical operation data. The equipment health status evaluation model is continuously updated and improved to improve the accuracy of fault prediction. The optimized equipment operating environment parameters are used as the new preset threshold range. When the real-time operation data of the equipment exceeds the threshold again, the above early warning, maintenance, and optimization process are repeated to continuously improve the equipment performance through closed-loop control. Regularly evaluate the effectiveness of the equipment maintenance strategy adopted. Continuously improve the maintenance strategy by comparing the equipment performance indicators and maintenance costs before and after optimization. If the frequency of equipment failure warnings increases significantly and the maintenance cost increases significantly, it is necessary to consider updating the equipment health status assessment model and adjusting the preset threshold range to adapt to the actual operating conditions of the equipment.

[0089] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. An intelligent operation and maintenance method for a solar thermal power generation system based on the Internet of Things, characterized in that: The method comprises: Obtain temperature, humidity, and vibration environmental parameter data collected by sensors deployed on power generation equipment; Receive environmental parameter data, compare the environmental parameter data with preset thresholds, and determine whether the environmental parameters exceed a normal range; When environmental parameters exceed preset thresholds, an early warning message containing the abnormal parameter type, value, and occurrence time is generated, and the priority is evaluated based on the early warning message; According to the priority of the early warning information, the power output, fuel consumption rate, and vibration frequency performance parameters of the power generation equipment are collected, and the health status of the power generation equipment is evaluated by combining the environmental parameter data and performance parameters using a weighted average algorithm to obtain a health index. Generate a maintenance plan when the health index of the power generation equipment is below the preset threshold; Track the progress of maintenance plan execution, obtain equipment performance parameters after maintenance, and use machine learning algorithms to update the health status model of power generation equipment based on performance parameters and environmental parameters, including adjusting the weight coefficients of each parameter; Based on the updated health status model, a dynamic threshold adjustment algorithm is used to automatically adjust the preset thresholds of environmental parameters and performance parameters to optimize the early warning mechanism and maintenance strategy.

2. The method according to claim 1, characterized in that The acquisition of temperature, humidity, and vibration environment parameter data collected by sensors deployed on the power generation equipment includes: Obtaining environmental parameter data collected by temperature, humidity and vibration sensors on the power generation equipment, and determining whether the data exceeds a preset threshold; If exceeded, the current operation data and historical operation data of the power generation equipment are obtained, an isolation forest algorithm is used to perform anomaly detection on the current operation data, and the detection result is compared with the historical operation data to determine the specific cause of the anomaly; According to the abnormal cause, relevant operating parameters are extracted as decision tree model features, and the decision tree model is trained to generate corresponding operation and maintenance strategies.

3. The method according to claim 1, characterized in that The receiving of environmental parameter data, comparing the environmental parameter data with a preset threshold, and determining whether the environmental parameter exceeds a normal range includes: Obtain the preprocessed environmental parameter data and preset thresholds, and use the cosine similarity algorithm to calculate the similarity score; If the similarity score is lower than the threshold, an abnormal alarm is triggered; Obtain the severity and duration of abnormal situations and generate emergency response plans through the CART decision tree algorithm; The exception handling process and result data are stored in the database, and the historical data is analyzed using the LSTM neural network algorithm to dynamically adjust the threshold settings and optimize the monitoring system.

4. The method according to claim 1, characterized in that When the environmental parameters exceed the preset threshold, an early warning message including the abnormal parameter type, value, and occurrence time is generated, and the priority is evaluated according to the early warning message, including: Acquire various environmental parameter data collected by the sensor in real time, and send the environmental parameter data to the data processing module through the wireless transmission module; Determine whether the environmental parameter data exceeds a preset threshold value, and if so, trigger a pre-established abnormality determination mechanism; According to the priority level of the warning information and in combination with a preset push strategy, a push method and a push channel for the warning information are determined, and the warning information is pushed to at least one target terminal.

5. The method according to claim 1, characterized in that According to the priority of the early warning information, the power output, fuel consumption rate, and vibration frequency performance parameters of the power generation equipment are collected, and the health status of the power generation equipment is evaluated by combining the environmental parameter data and the performance parameters using a weighted average algorithm to obtain a health index, including: Obtaining the weight coefficients of the preset performance parameters and environmental parameters; The weighted average algorithm is used to perform weighted calculation on the collected performance parameters and environmental parameters to obtain comprehensive evaluation indicators; Comparing the comprehensive evaluation index with a preset health status threshold to determine the health status level of the power generation equipment; According to the health status level, combined with a preset health index calculation formula, calculate the health index value of the power generation equipment; If the health index value is higher than a preset threshold, it is determined that the power generation equipment is in good operating condition; Otherwise, it is determined that there is a potential failure risk.

6. The method according to claim 1, characterized in that When the health index of the power generation equipment is lower than the preset threshold, a maintenance plan is generated, which includes maintenance items, required personnel, tools, spare parts and estimated time, including: Acquire real-time health index data of the power generation equipment and determine whether the health index is lower than a preset threshold; If the health index is lower than the threshold, a CART decision tree algorithm is used to generate a maintenance item list by taking equipment parameters as features and whether maintenance is required as the target variable and training based on historical maintenance data; According to the maintenance project list, obtain required resource information; By connecting with the ERP system, real-time inventory of resources can be obtained. If resources are insufficient, the priority of the maintenance plan can be adjusted according to the urgency of the maintenance project to optimize the feasibility of the maintenance plan.

7. The method according to claim 1, characterized in that The tracking of the execution progress of the maintenance plan, obtaining the performance parameters of the equipment after maintenance, and updating the health status model of the power generation equipment using a machine learning algorithm according to the performance parameters and environmental parameters, including adjusting the weight coefficient of each parameter, include: Obtain equipment performance data and environmental parameter data and input them into a pre-built equipment health status assessment model trained using a support vector machine algorithm; According to the difference between the health status value calculated by the model and the actual equipment operation status, a grid search method is used to adjust the penalty coefficient and kernel function type hyperparameters of the support vector machine model to optimize the model; The optimized equipment health status assessment model is deployed to the production environment, and the real-time collected equipment performance data and environmental parameter data are predicted to output the equipment health status prediction results.

8. The method according to claim 1, characterized in that The method of automatically adjusting preset thresholds of environmental parameters and performance parameters using a dynamic threshold adjustment algorithm based on the updated health status model to optimize the early warning mechanism and maintenance strategy includes: Obtain historical equipment operation data, determine the relationship between various performance parameters and equipment health status through correlation analysis, and establish an equipment health status assessment model; For the newly collected real-time operation data of the equipment, the support vector machine algorithm is used for incremental learning, which is added to the original historical operation data, so as to continuously update and improve the equipment health status assessment model. By comparing the equipment performance indicators and maintenance costs before and after optimization, the maintenance strategy is continuously improved.

9. An intelligent operation and maintenance system for a solar thermal power generation system based on the Internet of Things, based on the intelligent operation and maintenance method for a solar thermal power generation system based on the Internet of Things according to any one of claims 1 to 8, characterized in that: Environmental parameter acquisition module, environmental parameter monitoring and alarm module, equipment health status assessment module, maintenance plan generation and optimization module, model update and optimization module, anomaly detection and processing module and model update module; Environmental parameter acquisition module, used to obtain temperature, humidity, and vibration environmental parameter data collected by sensors deployed on power generation equipment; The environmental parameter monitoring and alarm module is used to receive environmental parameter data, compare the environmental parameter data with the preset threshold value, and determine whether the environmental parameter exceeds the normal range; The equipment health status assessment module is used to generate warning information containing abnormal parameter type, value, and occurrence time when environmental parameters exceed the preset threshold, and evaluate the priority according to the warning information; The maintenance plan generation and optimization module is used to trigger the collection of power output, fuel consumption rate, and vibration frequency performance parameters of power generation equipment according to the priority of early warning information, and use the weighted average algorithm to combine environmental parameter data and performance parameters to evaluate the health status of the power generation equipment and obtain the health index; Model update and optimization module, used to generate maintenance plans when the health index of power generation equipment falls below a preset threshold; Anomaly detection and processing module, used to track the progress of maintenance plan execution and obtain equipment performance parameters after maintenance; The model update module uses a dynamic threshold adjustment algorithm to automatically adjust the preset thresholds of environmental parameters and performance parameters according to the updated health status model, and optimizes the early warning mechanism and maintenance strategy.

10. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the intelligent operation and maintenance method of a solar thermal power generation system based on the Internet of Things according to any one of claims 1 to 8 are implemented.

Images