Real-time monitoring and diagnosis method for operation energy efficiency of biomass energy power station

Abstract

The invention relates to the technical field of energy efficiency monitoring of biomass energy power stations, and discloses a real-time monitoring and diagnosis method for operation energy efficiency of a biomass energy power station, which comprises the following steps of: acquiring fuel characteristics, boiler combustion, turbine operation and environment-friendly emission parameters in the operation process of the biomass energy power station through a multi-source heterogeneous sensor network; generating a multi-dimensional real-time monitoring data set; performing energy efficiency key index feature extraction based on the multi-dimensional real-time monitoring data set to generate an energy efficiency key index feature vector set; when the operation energy efficiency of the biomass energy power station is monitored in real time, a multi-source heterogeneous sensor network and an energy efficiency key index evaluation system are established to collect operation multi-dimensional parameters of the power station in real time and perform data fusion processing; the problem of key energy efficiency index feature distortion caused by sensor drift and communication interference in the data acquisition process can be identified in real time, the accuracy of energy efficiency state evaluation is ensured, and the energy efficiency monitoring error is reduced.

Description

Technical Field

[0001] This invention relates to the field of energy efficiency monitoring technology for biomass power plants, specifically a method for real-time monitoring and diagnosis of the operating energy efficiency of biomass power plants. Background Technology

[0002] Biomass power generation is a type of renewable energy generation that utilizes the biomass energy inherent in biomass. It includes power generation from the direct combustion of agricultural and forestry waste, power generation from the gasification of agricultural and forestry waste, power generation from waste incineration, power generation from landfill gas, and biogas power generation. China is a major agricultural country with abundant biomass resources. Various crops produce a huge amount of straw every year, and there is enormous potential for development through the utilization of straw for power generation.

[0003] Currently, in the monitoring and diagnosis of energy efficiency of biomass power plants, due to the complexity of the power plant system and frequent fluctuations in operating conditions, the multi-source operating parameters collected by traditional sensor networks cannot be used to identify in real time whether the key energy efficiency indicators are distorted due to sensor drift and communication interference during the data acquisition process. When data acquisition anomalies occur and are not identified in time, it will cause large deviations in energy efficiency status assessment and cannot guarantee the accuracy of energy efficiency monitoring results.

[0004] Therefore, a real-time monitoring and diagnosis method for the energy efficiency of biomass power plants is proposed to solve the above problems. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a real-time monitoring and diagnosis method for the energy efficiency of biomass power plants, solving the problem mentioned in the background technology of large deviations in energy efficiency status assessment and inability to guarantee the accuracy of energy efficiency monitoring results.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant, the method comprising the following steps: S1. Collect fuel characteristics, boiler combustion, turbine operation and environmental emission parameters during the operation of biomass power plants through a multi-source heterogeneous sensor network to generate a multi-dimensional real-time monitoring dataset; S2. Extract key energy efficiency indicators from a multi-dimensional real-time monitoring dataset and generate a set of key energy efficiency indicator feature vectors. S3. Based on the feature vector set of key energy efficiency indicators, conduct real-time energy efficiency status assessment and generate a real-time energy efficiency status assessment report. S4. Based on the real-time energy efficiency status assessment report, identify energy efficiency anomaly diagnosis patterns and generate energy efficiency anomaly diagnosis pattern feature data. S5. Based on the energy efficiency anomaly diagnosis pattern feature data and the historical energy efficiency anomaly case library, perform similarity matching to generate the root cause diagnosis results of energy efficiency anomalies. S6. Based on the root cause diagnosis results of energy efficiency anomalies, intelligently recommend energy efficiency optimization strategies and generate a set of operational energy efficiency optimization strategies. S7. Based on the root cause diagnosis results of energy efficiency anomalies and the set of energy efficiency optimization strategies, a real-time monitoring and diagnosis model for the operating energy efficiency of biomass power plants is constructed using digital twin technology. S8. Encrypt and store the energy efficiency status data and historical energy efficiency assessment reports output by the digital twin model using blockchain to establish a trusted storage and sharing mechanism for energy efficiency data. S9. Utilize artificial intelligence algorithms to predict and analyze energy efficiency trends, and generate an energy efficiency trend prediction report. S10. Integrate the energy efficiency prediction results to establish a comprehensive energy efficiency evaluation system for biomass power plants and generate a comprehensive energy efficiency evaluation index.

[0007] Preferably, the parameters acquired by the multi-source heterogeneous sensor network in S1 include the acquisition of a multi-dimensional real-time monitoring dataset. This dataset specifically includes fuel characteristic parameters, boiler combustion parameters, turbine operating parameters, and environmental emission parameters, achieved through the following steps: S11. Real-time data collection of biomass fuel's moisture content, calorific value, ash content, and volatile matter characteristics using an online fuel quality monitoring system, generating a real-time biomass fuel characteristic monitoring data matrix. ,in Indicates the number of monitoring time points. Indicates the dimension of fuel characteristic parameters; S12. Collect boiler temperature distribution, combustion efficiency, flue gas composition, and exhaust gas temperature parameters through the boiler combustion status monitoring system to generate a real-time boiler combustion status monitoring data matrix. ,in This indicates the number of boiler monitoring points. This indicates the dimension of the combustion state parameters; S13. Collect turbine speed, power output, thermal efficiency, and vibration characteristic parameters through the turbine performance monitoring system to generate a real-time monitoring data matrix of turbine operating status. ,in This indicates the number of monitoring periods for the steam turbine. Indicates the dimension of the running status parameters; S14. Collect environmental parameters such as flue gas emission concentration, pollutant content, and desulfurization and denitrification efficiency through the environmental monitoring system, and generate a real-time environmental emission monitoring data matrix. ,in Indicates the length of the emissions monitoring time series. This indicates the dimensions of environmental protection parameters.

[0008] Preferably, the feature extraction of key energy efficiency indicators in S2 includes the following steps: S21. Establish an evaluation system for key energy efficiency indicators of biomass power plants, including thermal efficiency indicators. Power generation efficiency indicators Comprehensive energy utilization rate index Environmental performance indicators ; S22. Calculate energy efficiency indicators based on multi-dimensional real-time monitoring datasets, and use data fusion algorithms to eliminate sensor measurement errors and generate energy efficiency indicator datasets. S23. Principal component analysis is used to reduce the dimensionality of key energy efficiency indicators, extracting the feature vectors that best reflect the operating status of biomass power plants, and generating a feature vector set of key energy efficiency indicators. .

[0009] Preferably, the real-time energy efficiency status assessment in S3 includes the following steps: S31. Construct an energy efficiency status assessment index system for biomass power plants, including operational economic indicators. Technical indicators Reliability indicators and environmental indicators ; S32. Use the fuzzy comprehensive evaluation method to assign weights and normalize the various energy efficiency indicators to generate a standardized energy efficiency evaluation matrix. S33. The objective weights of each energy efficiency indicator are determined based on the entropy weight method, the subjective weights are determined by combining the expert scoring method, and the comprehensive weight vector is generated by the combined weighting method. S34. Calculate the real-time comprehensive energy efficiency score of the biomass power plant using a weighted average algorithm, and generate an energy efficiency status assessment report. .

[0010] Preferably, the energy efficiency anomaly diagnostic pattern recognition in S4 includes the following steps: S41. Establish a knowledge base for diagnosing energy efficiency anomalies in biomass power plants, including feature vector sets of typical energy efficiency anomaly patterns. ,in Indicates the number of exception patterns; S42. Use deep learning algorithms to analyze real-time energy efficiency data and identify abnormal fluctuation patterns and trend change characteristics; S43. Detect abnormal energy efficiency data points using the isolated forest algorithm, and identify the duration and development trend of the abnormality by combining time series analysis; S44. Classify energy efficiency anomaly patterns based on cluster analysis and generate characteristic data for energy efficiency anomaly diagnosis patterns. .

[0011] Preferably, the similarity matching analysis in S5 includes the following steps: S51. Construct a historical energy efficiency anomaly case library Each case study includes information on anomaly characteristics, root cause analysis, treatment measures, and effectiveness evaluation. Number of cases; S52. Calculate the similarity between the current energy efficiency anomaly pattern and historical cases using the cosine similarity algorithm, and generate a similarity matching matrix. ; S53. Based on the K-nearest neighbor algorithm, find the historical cases most similar to the current anomaly and analyze the root cause of the anomaly. S54. Analyze the causal relationship between energy efficiency anomalies and operating parameters using association rule mining techniques to generate an energy efficiency anomaly root cause diagnosis report. .

[0012] Preferably, the intelligent recommendation of energy efficiency optimization strategies in S6 includes the following steps: S61. Establish a knowledge base for energy efficiency optimization strategies for biomass power plants, including equipment adjustment strategies. Optimization strategies Maintenance and management strategies ; S62. Based on the case-based reasoning technology, recommend appropriate optimization strategies according to the energy efficiency anomaly diagnosis results, and generate a preliminary set of optimization schemes; S63. Employ a multi-objective optimization algorithm to simulate the effectiveness and evaluate the benefits of the recommendation strategy, and determine the optimal energy efficiency improvement path; S64. Prioritize the various optimization strategies using the Analytic Hierarchy Process (AHP) to generate a set of energy efficiency optimization strategies. .

[0013] Preferably, the digital twin technology model construction in S7 includes the following steps: S71. Construct a digital twin model of a biomass power plant based on physical modeling and data-driven methods. The model includes a virtual-real mapping module and a dynamic simulation module. S72. The energy efficiency optimization scheme is virtually verified using a digital twin model. The verification process includes scheme effect simulation and risk identification operations. S73. Establish a real-time interaction mechanism between energy efficiency monitoring data and digital twin model, wherein the mechanism supports the visualization of energy efficiency status and maintenance early warning; S74. Iterative optimization of energy efficiency optimization strategy using digital twin technology, wherein the optimization process includes strategy parameter adjustment and model version update.

[0014] Preferably, the blockchain technology used to establish the storage mechanism in S8 includes the following steps: S81. Construct a distributed storage network for energy efficiency data of biomass power plants based on blockchain. The network consists of multiple nodes, each of which stores a complete copy of the energy efficiency data and synchronizes the data status through a consensus mechanism. S82. Smart contract technology is used to automatically verify and share energy efficiency data. The smart contract includes data format verification rules, access control logic, and data transaction execution conditions. S83. Apply timestamps and hash algorithms to process energy efficiency monitoring data, generate a data hash sequence containing timestamps, and link each data block through a chain structure; S84. Establish a blockchain-based energy efficiency data traceability system, which records the entire chain of energy efficiency data operation logs from collection and storage to access, and supports forward and reverse tracing based on data hash.

[0015] Preferably, the energy efficiency trend prediction using artificial intelligence algorithms in step S9 includes the following steps: S91. Use the LSTM neural network algorithm to predict the trend of energy efficiency of biomass power plants and generate short-term, medium-term and long-term energy efficiency prediction reports. S92. Establish an energy efficiency early warning mechanism based on the prediction results, and set multi-level early warning thresholds and corresponding emergency response plans.

[0016] Compared with existing technologies, this invention provides a method for real-time monitoring and diagnosis of the energy efficiency of biomass power plants, which has the following beneficial effects: 1. In this invention, when conducting real-time monitoring of the energy efficiency of a biomass power plant, a multi-source heterogeneous sensor network and an evaluation system for key energy efficiency indicators are established. Multi-dimensional parameters of the power plant operation are collected in real time and data fusion processing is performed. This enables real-time identification of whether the data collection process is affected by sensor drift and communication interference, which leads to distortion of key energy efficiency indicators. This ensures the accuracy of energy efficiency status assessment and reduces energy efficiency monitoring errors.

[0017] 2. In this invention, when diagnosing energy efficiency anomalies in biomass power plants, an energy efficiency anomaly diagnosis knowledge base is constructed and intelligent algorithms are used for pattern recognition and similarity matching. This verifies in real time whether the energy efficiency anomaly pattern recognition algorithm processes multi-dimensional feature data abnormally, thereby reducing the possibility of misjudging the root cause of energy efficiency anomalies. Furthermore, when diagnosing logical anomalies, case reasoning technology can be used to correct them in real time, ensuring the accuracy of energy efficiency anomaly diagnosis.

[0018] 3. In this invention, when optimizing the energy efficiency of biomass power plant operation, a set of energy efficiency optimization strategies matching dynamic operating conditions is generated in real time through a multi-objective optimization algorithm and an adaptive strategy generation mechanism. This enables multi-strategy collaborative optimization, reduces the situation where the energy efficiency improvement scheme is not well matched with the actual operating conditions, and further improves the stability and economy of energy efficiency optimization of biomass power plant operation. Attached Figure Description

[0019] Figure 1 This is a flowchart of a method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] For specific implementation examples, please refer to: Figure 1 A method for real-time monitoring and diagnosis of the energy efficiency of a biomass power plant, comprising the following steps: S1. Collect fuel characteristics, boiler combustion, turbine operation and environmental emission parameters during the operation of biomass power plants through a multi-source heterogeneous sensor network to generate a multi-dimensional real-time monitoring dataset; S2. Extract key energy efficiency indicators from a multi-dimensional real-time monitoring dataset and generate a set of key energy efficiency indicator feature vectors. S3. Based on the feature vector set of key energy efficiency indicators, conduct real-time energy efficiency status assessment and generate a real-time energy efficiency status assessment report. S4. Based on the real-time energy efficiency status assessment report, identify energy efficiency anomaly diagnosis patterns and generate energy efficiency anomaly diagnosis pattern feature data. S5. Based on the energy efficiency anomaly diagnosis pattern feature data and the historical energy efficiency anomaly case library, perform similarity matching to generate the root cause diagnosis results of energy efficiency anomalies. S6. Based on the root cause diagnosis results of energy efficiency anomalies, intelligently recommend energy efficiency optimization strategies and generate a set of operational energy efficiency optimization strategies. S7. Based on the root cause diagnosis results of energy efficiency anomalies and the set of energy efficiency optimization strategies, a real-time monitoring and diagnosis model for the operating energy efficiency of biomass power plants is constructed using digital twin technology. S8. Encrypt and store the energy efficiency status data and historical energy efficiency assessment reports output by the digital twin model using blockchain to establish a trusted storage and sharing mechanism for energy efficiency data. S9. Utilize artificial intelligence algorithms to predict and analyze energy efficiency trends, and generate an energy efficiency trend prediction report. S10. Integrate energy efficiency prediction results, establish a comprehensive evaluation system for the energy efficiency of biomass power plants, and generate a comprehensive energy efficiency evaluation index.

[0022] The parameters acquired by the multi-source heterogeneous sensor network in S1 include the acquisition of multi-dimensional real-time monitoring datasets. These datasets specifically include fuel characteristic parameters, boiler combustion parameters, turbine operating parameters, and environmental emission parameters. This is achieved through the following steps: S11. Real-time data collection of biomass fuel's moisture content, calorific value, ash content, and volatile matter characteristics using an online fuel quality monitoring system, generating a real-time biomass fuel characteristic monitoring data matrix. ,in Indicates the number of monitoring time points. Indicates the dimension of fuel characteristic parameters; S12. Collect boiler temperature distribution, combustion efficiency, flue gas composition, and exhaust gas temperature parameters through the boiler combustion status monitoring system to generate a real-time boiler combustion status monitoring data matrix. ,in This indicates the number of boiler monitoring points. This indicates the dimension of the combustion state parameters; S13. Collect turbine speed, power output, thermal efficiency, and vibration characteristic parameters through the turbine performance monitoring system to generate a real-time monitoring data matrix of turbine operating status. ,in This indicates the number of monitoring periods for the steam turbine. Indicates the dimension of the running status parameters; S14. Collect environmental parameters such as flue gas emission concentration, pollutant content, and desulfurization and denitrification efficiency through the environmental monitoring system, and generate a real-time environmental emission monitoring data matrix. ,in Indicates the length of the emissions monitoring time series. This indicates the dimensions of environmental protection parameters.

[0023] The feature extraction of key energy efficiency indicators in S2 includes the following steps: S21. Establish an evaluation system for key energy efficiency indicators of biomass power plants, including thermal efficiency indicators. Power generation efficiency indicators Comprehensive energy utilization rate index Environmental performance indicators ; S22. Calculate energy efficiency indicators based on a multi-dimensional real-time monitoring dataset, and use a data fusion algorithm to eliminate sensor measurement errors to generate an energy efficiency indicator dataset. This includes the following steps: S221. Obtain a multi-dimensional real-time monitoring dataset, including a real-time monitoring data matrix of fuel characteristics. Boiler Combustion Status Real-time Monitoring Data Matrix Real-time monitoring data matrix of steam turbine operating status Real-time monitoring data matrix of environmental protection emissions ; S222. Perform data standardization on the multi-dimensional real-time monitoring dataset. Use the Z-score standardization method to eliminate the differences in the dimensions of different sensors and generate a standardized monitoring data matrix. The standardization process includes calculating the mean and standard deviation of each data point and performing a linear transformation. S223. A weighted average fusion algorithm is used to fuse the standardized monitoring data and calculate the fusion value of each energy efficiency index. The fusion algorithm is based on the sensor weight coefficient and the standardized data to perform a weighted average calculation. S224. Calculate the energy efficiency index dataset based on the fused data, including thermal efficiency index. Power generation efficiency indicators Comprehensive energy utilization rate index and environmental performance indicators ; S225. Perform outlier detection and correction on the energy efficiency index dataset, use statistical methods to identify and eliminate sensor measurement errors, and generate an accurate energy efficiency index dataset. S23. Principal component analysis is used to reduce the dimensionality of key energy efficiency indicators, extracting the feature vectors that best reflect the operating status of biomass power plants, and generating a feature vector set of key energy efficiency indicators. .

[0024] The real-time energy efficiency status assessment in S3 includes the following steps: S31. Construct an energy efficiency status assessment index system for biomass power plants, including operational economic indicators. Technical indicators Reliability indicators and environmental indicators ; S32. The fuzzy comprehensive evaluation method is used to assign weights and normalize the various energy efficiency indicators to generate a standardized energy efficiency evaluation matrix. This includes the following steps: S321. Establish a comment set for energy efficiency indicators. ,in Indicates the number of rating levels; S322. Constructing a fuzzy relation matrix for energy efficiency indicators. Matrix elements Indicates the first The indicator belongs to the first Membership degree of each rating level; S323, Combining weight vectors Calculate the fuzzy comprehensive evaluation results: ; in For fuzzy synthesis operators; S324. The evaluation results are defuzzified using the maximum membership principle to generate a standardized energy efficiency assessment matrix; S33. Based on the entropy weight method, determine the objective weights of each energy efficiency indicator, combine the expert scoring method to determine the subjective weights, and use the combined weighting method to generate a comprehensive weight vector. The specific implementation of the entropy weight method includes the following steps: S331. Normalize the energy efficiency index data to generate a normalized matrix, where the normalization formula is: ; in Indicates the first The first sample The normalized value of the indicator, These are the original index values. For sample index, For indexing indicators; S332, Calculate the first The entropy value of the item index is calculated using the following formula: ; in The normalization constant is Let be the entropy value of the j-th index; S333. Calculate the index difference coefficient based on the entropy value and generate the entropy weight: ; ; in Let be the difference coefficient of the j-th indicator. Let the entropy weight be the weight of the j-th index. For the number of indicators; S334. Combine the entropy weight with the subjective weight obtained by the expert scoring method to generate a comprehensive weight vector; The specific implementation of the expert scoring method includes the following steps: S335. Form a scoring group consisting of 5-10 experts in the field, design an energy efficiency index weighting scoring table, and score the range from 1 to 9 points. S336. Collect independent scores from experts and calculate the average score for each indicator. and standard deviation ; S337. Remove outlier scores that deviate from the mean ±2σ, take the arithmetic mean of the valid scores, and generate a subjective weight vector. S338. If the consistency test is passed, the scoring result is considered reliable; otherwise, the scoring should be reorganized. S34. Calculate the real-time comprehensive energy efficiency score of the biomass power plant using a weighted average algorithm, and generate an energy efficiency status assessment report. Specifically, it includes the following steps: S341. Calculate the real-time energy efficiency comprehensive score using the weighted average formula: ; in For real-time comprehensive energy efficiency score, Indicates the first The weight of each energy efficiency indicator, Indicates the first Standardized values ​​for each energy efficiency indicator. This represents the total number of energy efficiency indicators. S342. Compare the calculated real-time comprehensive energy efficiency score with the preset energy efficiency level threshold to determine the energy efficiency status level. S343. Generate an energy efficiency status assessment report based on the energy efficiency status level and comprehensive score. The report includes the energy efficiency score, status level, and improvement suggestions.

[0025] The energy efficiency anomaly diagnostic pattern identification in S4 includes the following steps: S41. Establish a knowledge base for diagnosing energy efficiency anomalies in biomass power plants, including feature vector sets of typical energy efficiency anomaly patterns. ,in Indicates the number of exception patterns; S42. Analyze real-time energy efficiency data using deep learning algorithms to identify abnormal fluctuation patterns and trend changes, specifically including the following steps: S421. Construct a convolutional neural network model. The input layer receives energy efficiency time series data. The convolutional layer is set with 3 convolutional kernels, and the pooling layer uses max pooling. S422. Using the ReLU activation function for nonlinear transformation, the fully connected layer outputs abnormal fluctuation pattern characteristics. S423. The model is trained using the stochastic gradient descent algorithm with a learning rate of 0.001 and more than 1000 iterations. S424. Evaluate the model accuracy through cross-validation, with the accuracy threshold set at 95%. The specific implementation of the stochastic gradient descent algorithm includes the following steps: S4231. Initialize model parameters, including convolutional layer weight matrices. Fully connected layer weight matrix and bias vector The initial weights are randomly sampled from a normal distribution with a mean of 0 and a standard deviation of 0.01, and the initial bias is 0. S4232. Randomly select a mini-batch of samples from the energy efficiency training dataset, setting the batch size to 32, and calculate the loss function value of the current batch. The loss function used is cross-entropy loss:

[0026] in The value of the loss function. For batch size, For the number of categories, For category indexing, For sample index, For real labels, To predict probabilities; S4233. Calculate the gradient of the loss function with respect to the model parameters, using the backpropagation algorithm to calculate the gradient values ​​layer by layer: ; in , These represent the weight matrix and the bias vector, respectively. Represents the loss function For the weight matrix gradient, Represents the loss function For bias vector The gradient; S4234. Update model parameters using the following formula: ; in , These are the updated weight matrix and bias vector, respectively. , The weight matrix and bias vector before the update, and the learning rate. The updated parameters will be used for the next iteration; S4235. Repeat steps S4232 to S4234 until the number of iterations reaches 1000 and the rate of change of the loss function is less than 1000. ; S43. Detect abnormal energy efficiency data points using the isolated forest algorithm, and identify the duration and development trend of the anomalies by combining time series analysis. This includes the following steps: S431. Randomly sample from the energy efficiency dataset to generate multiple isolated trees, with the maximum depth of each tree set to... ,in The number of samples; S432. Calculate the path length from the root node to the leaf node for each data point, and generate a set of path lengths; S433. Calculate the anomaly score based on the path length, using the following formula: ; in Indicates abnormal scores. For data points Average path length, For standardization factors, The number of samples; S434. Set the abnormal score threshold to 0.6 to identify abnormal energy efficiency data points; S44. Classify energy efficiency anomaly patterns based on cluster analysis and generate characteristic data for energy efficiency anomaly diagnosis patterns. Specifically, it includes the following steps: S441. The K-means algorithm is used to cluster the energy efficiency anomaly patterns. The number of cluster centers is initialized to k=3, corresponding to mild, moderate and severe anomalies. S442. Calculate the similarity between data points and cluster centers using distance metrics, and assign data points to corresponding clusters based on the nearest neighbor principle; S443. Iterate and update the cluster center until the change in the cluster center is less than 0.001 and the number of iterations reaches 100. S444. Output clustering labels to generate energy efficiency anomaly diagnostic pattern feature data. ; S4441. For each data point in the energy efficiency anomaly pattern dataset and cluster center Calculate their Euclidean distance: ; in Representing data points With cluster center distance, Indicates the first The first data point 3D eigenvalues Indicates the first The first cluster center 3D coordinates For feature dimension, For indexing data points, For cluster center index, Indexed by feature dimensions; S4442. Data points are assigned to corresponding clusters according to the nearest neighbor principle. The cluster label assignment formula is as follows: ; in Representing data points Cluster tags, For indexing data points, For cluster center index; S4443, Repeat S4441 to S4442 until all data points have completed cluster allocation.

[0027] Similarity matching analysis in S5 includes the following steps: S51. Construct a historical energy efficiency anomaly case library Each case study includes information on anomaly characteristics, root cause analysis, treatment measures, and effectiveness evaluation. Number of cases; S52. Calculate the similarity between the current energy efficiency anomaly pattern and historical cases using the cosine similarity algorithm, and generate a similarity matching matrix. Specifically, it includes the following steps: S521, from Extract the feature vector of the current energy efficiency anomaly mode ,from Extracting historical case feature vectors And represent both as points in a high-dimensional space; S522. Calculate the cosine of the angle between vectors using the following formula: ; in Indicates similarity. Represents the vector dot product. and Indicates the magnitude of the vector; S523. Map the cosine value to the [0,1] interval to generate a similarity matching matrix. ; S524. Set a similarity threshold of 0.7 to filter historical cases with high similarity. S53. Based on the K-nearest neighbor algorithm, find the historical cases most similar to the current anomaly, and analyze the root cause of the anomaly. This includes the following steps: S531. Based on historical energy efficiency data, the nearest neighbor number K is determined through cross-validation optimization, and the feature vector of the current abnormal mode is calculated using the Euclidean distance formula. Feature vectors of historical cases Distance: ; in For feature dimension, Indexed by feature dimensions, The first feature vector representing the current anomaly pattern. 3D eigenvalues The first feature vector representing a historical case 3D eigenvalues Indicates Euclidean distance; S532, Select the one with the smallest distance. Each case is used as a nearest neighbor set; S533. Based on the labels of neighboring cases, vote to generate a preliminary diagnosis of the root cause of the anomaly; S534. If the voting results in a tie, then add... Value to Recalculate; S54. Analyze the causal relationship between energy efficiency anomalies and operating parameters using association rule mining techniques to generate an energy efficiency anomaly root cause diagnosis report. Specifically, it includes the following steps: S541. Construct a trading database of energy efficiency anomalies and operating parameters, with each record containing an anomaly identifier and the corresponding combination of operating parameters; S542. Frequent itemsets are generated using the Apriori algorithm, and the minimum support threshold is determined through lift metric optimization. ; in To increase the degree, Represents the support of an itemset, when The time indicates that the rule has a positive correlation; S543. Extract association rules from the frequent itemset, optimize and determine the minimum confidence threshold based on the dual indicators of lift and confidence, and generate association rules that simultaneously satisfy high confidence and statistical significance. S544. Output strong correlation rules to generate an energy efficiency anomaly root cause diagnosis report. .

[0028] The intelligent recommendation of energy efficiency optimization strategies in S6 includes the following steps: S61. Establish a knowledge base for energy efficiency optimization strategies for biomass power plants, including equipment adjustment strategies. Optimization strategies Maintenance and management strategies ; S62. Based on the energy efficiency anomaly diagnosis results, the case-based reasoning technology recommends suitable optimization strategies and generates a preliminary set of optimization solutions, which includes the following steps: S621. Search the historical energy efficiency anomaly case database. The cosine similarity algorithm is used to match the current case; S622. Adapt and modify the optimization strategy for the matching cases to generate a preliminary optimization plan; S623. Verify the feasibility of the solution through the rule engine and eliminate conflicting strategies; S624. Output the set of optimized solutions for further processing by multi-objective optimization algorithms; S63. Employ a multi-objective optimization algorithm to simulate the effectiveness and evaluate the benefits of the recommendation strategy, and determine the optimal energy efficiency improvement path. The specific implementation of the multi-objective optimization algorithm includes the following steps: S631. Establish the optimization objective function, including the energy efficiency improvement rate. Cost reduction rate and the rate of improvement of environmental protection indicators ,in: ; ; ; in Indicates energy efficiency. For the new efficiency value, As the baseline efficiency value, Indicates cost, Based on the baseline cost value, For the new cost value, Indicates environmental protection indicators, For the new energy value, This is the baseline energy value; S632. The NSGA-II algorithm is used for multi-objective optimization. The algorithm parameters, including population size, crossover probability and mutation probability, are determined through hyperparameter optimization techniques. S633. Generate the Pareto optimal solution set and select the optimal solution using the crowding ranking algorithm. The crowding degree calculation adopts the density estimation method of the solutions around the individual. The crowding degree calculation adopts the following steps: S6331. Sort the Pareto optimal solution set for each objective function to obtain an ordered solution sequence; S6332, For each solution Calculate its value in each objective function The congestion distance value above: ; in To solve In the The crowding distance portion of the objective function, Solution In the The value of the objective function and They represent the first The maximum and minimum values ​​of the objective function. To deindex, Index for the objective function; S6333. Summing the partial values ​​of each solution across all objective functions yields the total crowding distance: ; in To solve Total congestion distance, The total number of objective functions. To deindex, Index of the objective function, The starting index for summation; S6334, Distance based on congestion level Sort the solutions from largest to smallest and select the one with the highest crowding as the optimal solution. S634, Output the optimal energy efficiency improvement path; S64. Prioritize the various optimization strategies using the Analytic Hierarchy Process (AHP) to generate a set of energy efficiency optimization strategies. Specifically, it includes the following steps: S641. Constructing a strategy priority judgment matrix The importance of each energy efficiency optimization strategy was compared pairwise using the 1-9 scaling method, and the matrix elements were analyzed. Indicates the first The strategy relative to the first The importance of each strategy, among which and All are strategy serial numbers; S642, Calculate the judgment matrix Maximum eigenvalue and its corresponding eigenvectors , where the feature vector The calculation uses the sum-product method, specifically including: Judgment Matrix Normalize by column: ; Where α, β, γ, and ν are index variables. The order of the matrix; Sum the normalized matrix row by row: ; Normalizing the rows and vectors yields the weight vector: ; in This represents the weight of the γth policy term. ; S643. To perform a consistency check, first calculate the consistency index: ; in As a consistency indicator, To determine the largest eigenvalue of a matrix, The order of the matrix; Then query the average random consistency index. The table retrieves the corresponding... value; Finally, calculate the consistency ratio: ; in The consistency ratio, As the average random consistency index, when The consistency of the judgment matrix is ​​considered acceptable at that time. S644, The feature vector Normalization is performed to obtain the final weights of each strategy, and the strategies are then prioritized from largest to smallest to form a set of energy efficiency optimization strategies. .

[0029] The digital twin technology model in S7 includes the following steps: S71. Construct a digital twin model of a biomass power plant based on physical modeling and data-driven methods. The model includes a virtual-real mapping module and a dynamic simulation module. S72. Virtually verify the energy efficiency optimization scheme through a digital twin model. The verification process includes scheme effect simulation and risk identification. S73. Establish a real-time interaction mechanism between energy efficiency monitoring data and digital twin models, which supports the visualization of energy efficiency status and maintenance early warning. S74. Utilize digital twin technology to iteratively optimize energy efficiency optimization strategies. The optimization process includes adjusting strategy parameters and updating model versions.

[0030] The steps involved in establishing a storage mechanism using blockchain technology in S8 are as follows: S81. Construct a distributed storage network for energy efficiency data of biomass power plants based on blockchain. The network consists of multiple nodes, each of which stores a complete copy of the energy efficiency data and synchronizes the data status through a consensus mechanism. S82. Use smart contract technology to automatically verify and share energy efficiency data. The smart contract includes data format verification rules, access control logic and data transaction execution conditions. S83. Apply timestamps and hash algorithms to process energy efficiency monitoring data, generate a data hash sequence containing timestamps, and link each data block through a chain structure; S84. Establish a blockchain-based energy efficiency data traceability system. The system records the entire chain of energy efficiency data operation logs from collection and storage to access, and supports forward and reverse tracing based on data hash.

[0031] The S9 uses artificial intelligence algorithms to predict energy efficiency trends, including the following steps: S91. Using the LSTM neural network algorithm, trend prediction of the energy efficiency of biomass power plants is performed, generating short-term, medium-term, and long-term energy efficiency prediction reports. This includes the following steps: S911. Construct an LSTM network model. The number of nodes in the input layer of the model is consistent with the dimension of the feature vector of the key energy efficiency index in weight 2. The hidden layer adopts a multi-layer structure, and each hidden layer contains several neurons. S912. Organize historical energy efficiency data according to time series, generate training samples using the sliding window method, input them into the model, and output energy efficiency trend prediction values. Specifically, this includes: S9121, Set the length of the sliding window and sliding step size ,in This indicates the number of historical time points contained in each training sample. Indicates the number of time intervals during window sliding; S9122, Extracting continuous data from historical energy efficiency data sequences Data from each time point is used as input features. ,in Indicates the first Energy efficiency data at various time points; S9123, after the window Data from each time point was used as the prediction target. ,in Indicates the prediction time step; S9124, Using a sliding window in steps Traverse the entire historical dataset to generate a training sample set. ,in The total number of samples; S913. Update the model weights through the training algorithm to minimize the prediction error. The number of training iterations is determined based on the performance of the validation set. S914. Use the trained model to predict real-time energy efficiency data and generate an energy efficiency trend prediction report. S92. Establish an energy efficiency early warning mechanism based on the prediction results, and set multi-level early warning thresholds and corresponding emergency response plans.

[0032] The operation steps of this method for real-time monitoring and diagnosis of the energy efficiency of a biomass power plant are as follows: Step 1: Real-time acquisition of multi-source operational data By deploying a multi-source heterogeneous sensor network in key components of biomass power plants, fuel characteristic parameters, boiler combustion status parameters, turbine operating parameters, and environmental emission parameters are collected in real time. After frequency consistency judgment and sampling frequency adjustment, the collected raw data is processed to generate a standardized multi-dimensional real-time monitoring dataset, providing a complete data foundation for subsequent energy efficiency analysis.

[0033] Step 2: Extraction of Key Energy Efficiency Indicators Based on multi-dimensional real-time monitoring datasets, key energy efficiency indicators are extracted, and an energy efficiency evaluation system including thermal efficiency, power generation efficiency, comprehensive energy utilization rate, and environmental performance is established. Sensor measurement errors are eliminated through data fusion algorithms to generate accurate energy efficiency indicator datasets. Principal component analysis is used to reduce the dimensionality of key energy efficiency indicators and extract the feature vectors that best reflect the operating status of the power plant.

[0034] Step 3: Real-time assessment and analysis of energy efficiency status Based on the feature vector set of key energy efficiency indicators, real-time energy efficiency status assessment is conducted. An energy efficiency status assessment index system that includes operational economy, technology, reliability, and environmental protection is constructed. Fuzzy comprehensive evaluation method and entropy weight method are used to assign weights and normalize each energy efficiency indicator. The weighted average algorithm is used to calculate the real-time comprehensive energy efficiency score of the power plant.

[0035] Step 4: Intelligent Diagnosis and Identification of Energy Efficiency Anomalies An energy efficiency anomaly diagnosis knowledge base was established, and deep learning algorithms, isolated forest algorithms, and cluster analysis were used to diagnose energy efficiency anomalies. Real-time energy efficiency data was analyzed to identify abnormal fluctuation patterns and trend change characteristics, detect abnormal energy efficiency data points, and combine time series analysis to identify the duration and development trend of anomalies.

[0036] Step 5: Anomaly Root Cause Matching Diagnostic Analysis A historical energy efficiency anomaly case library is constructed. The cosine similarity algorithm is used to calculate the similarity between the current energy efficiency anomaly pattern and the historical cases. The K nearest neighbor algorithm is used to find the historical cases most similar to the current anomaly. The causal relationship between energy efficiency anomalies and operating parameters is analyzed through association rule mining technology.

[0037] Step Six: Intelligent Generation of Energy Efficiency Optimization Strategies Establish an energy efficiency optimization strategy knowledge base, recommend appropriate optimization strategies based on the energy efficiency anomaly diagnosis results using case reasoning technology, simulate the effects and evaluate the benefits of the recommended strategies using a multi-objective optimization algorithm, and prioritize the various optimization strategies using the AHP (Analytic Hierarchy Process).

[0038] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0039] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for real-time monitoring and diagnosis of operational energy efficiency of a biomass power plant, characterized in that: The method includes the following steps: S1. Collect fuel characteristics, boiler combustion, turbine operation and environmental emission parameters during the operation of biomass power plants through a multi-source heterogeneous sensor network to generate a multi-dimensional real-time monitoring dataset; S2. Extract key energy efficiency indicators from a multi-dimensional real-time monitoring dataset and generate a set of key energy efficiency indicator feature vectors. S3. Based on the feature vector set of key energy efficiency indicators, conduct real-time energy efficiency status assessment and generate a real-time energy efficiency status assessment report. S4. Based on the real-time energy efficiency status assessment report, identify energy efficiency anomaly diagnosis patterns and generate energy efficiency anomaly diagnosis pattern feature data. S5. Based on the energy efficiency anomaly diagnosis pattern feature data and the historical energy efficiency anomaly case library, perform similarity matching to generate the root cause diagnosis results of energy efficiency anomalies. S6. Based on the root cause diagnosis results of energy efficiency anomalies, intelligently recommend energy efficiency optimization strategies and generate a set of operational energy efficiency optimization strategies. S7. Based on the root cause diagnosis results of energy efficiency anomalies and the set of energy efficiency optimization strategies, a real-time monitoring and diagnosis model for the operating energy efficiency of biomass power plants is constructed using digital twin technology. S8. Encrypt and store the energy efficiency status data and historical energy efficiency assessment reports output by the digital twin model using blockchain to establish a trusted storage and sharing mechanism for energy efficiency data. S9. Utilize artificial intelligence algorithms to predict and analyze energy efficiency trends, and generate an energy efficiency trend prediction report. S10. Integrate the energy efficiency prediction results to establish a comprehensive energy efficiency evaluation system for biomass power plants and generate a comprehensive energy efficiency evaluation index.

2. The method of claim 1, wherein the method further comprises: The parameters acquired by the multi-source heterogeneous sensor network in S1 include the acquisition of multi-dimensional real-time monitoring datasets. These datasets specifically include fuel characteristic parameters, boiler combustion parameters, turbine operating parameters, and environmental emission parameters, and are achieved through the following steps: S11, real-time acquisition of the moisture content, heat value, ash content, and volatile matter characteristic parameters of the biomass fuel by the fuel quality online monitoring system to generate a biomass fuel characteristic real-time monitoring data matrix wherein represents the number of monitoring time points, represents the dimension of the fuel characteristic parameters; S12, collecting the boiler temperature distribution, combustion efficiency, flue gas composition, and exhaust gas temperature parameters through the boiler combustion state monitoring system to generate a boiler combustion state real-time monitoring data matrix wherein represents the number of boiler monitoring points, represents the combustion state parameter dimension; S13. Collect turbine speed, power output, thermal efficiency, and vibration characteristic parameters through the turbine performance monitoring system to generate a real-time monitoring data matrix of turbine operating status. ,in This indicates the number of monitoring periods for the steam turbine. Indicates the dimension of the running status parameters; S14. Collect environmental parameters such as flue gas emission concentration, pollutant content, and desulfurization and denitrification efficiency through the environmental monitoring system, and generate a real-time environmental emission monitoring data matrix. ,in Indicates the length of the emissions monitoring time series. This indicates the dimensions of environmental protection parameters.

3. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The feature extraction of key energy efficiency indicators in S2 includes the following steps: S21. Establish an evaluation system for key energy efficiency indicators of biomass power plants, including thermal efficiency indicators. Power generation efficiency indicators Comprehensive energy utilization rate index Environmental performance indicators ; S22. Calculate energy efficiency indicators based on multi-dimensional real-time monitoring datasets, and use data fusion algorithms to eliminate sensor measurement errors and generate energy efficiency indicator datasets. S23. Principal component analysis is used to reduce the dimensionality of key energy efficiency indicators, extracting the feature vectors that best reflect the operating status of biomass power plants, and generating a feature vector set of key energy efficiency indicators. .

4. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The real-time energy efficiency status assessment in S3 includes the following steps: S31. Construct an energy efficiency status assessment index system for biomass power plants, including operational economic indicators. Technical indicators Reliability indicators and environmental indicators ; S32. Use the fuzzy comprehensive evaluation method to assign weights and normalize the various energy efficiency indicators to generate a standardized energy efficiency evaluation matrix. S33. The objective weights of each energy efficiency indicator are determined based on the entropy weight method, the subjective weights are determined by combining the expert scoring method, and the comprehensive weight vector is generated by the combined weighting method. S34. Calculate the real-time comprehensive energy efficiency score of the biomass power plant using a weighted average algorithm, and generate an energy efficiency status assessment report. .

5. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The energy efficiency anomaly diagnostic pattern identification in S4 includes the following steps: S41. Establish a knowledge base for diagnosing energy efficiency anomalies in biomass power plants, including feature vector sets of typical energy efficiency anomaly patterns. ,in Indicates the number of exception patterns; S42. Use deep learning algorithms to analyze real-time energy efficiency data and identify abnormal fluctuation patterns and trend change characteristics; S43. Detect abnormal energy efficiency data points using the isolated forest algorithm, and identify the duration and development trend of the abnormality by combining time series analysis; S44. Classify energy efficiency anomaly patterns based on cluster analysis and generate characteristic data for energy efficiency anomaly diagnosis patterns. .

6. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The similarity matching analysis in S5 includes the following steps: S51. Construct a historical energy efficiency anomaly case library Each case study includes information on anomaly characteristics, root cause analysis, treatment measures, and effectiveness evaluation. Number of cases; S52. Calculate the similarity between the current energy efficiency anomaly pattern and historical cases using the cosine similarity algorithm, and generate a similarity matching matrix. ; S53. Based on the K-nearest neighbor algorithm, find the historical cases most similar to the current anomaly and analyze the root cause of the anomaly. S54. Analyze the causal relationship between energy efficiency anomalies and operating parameters using association rule mining techniques to generate an energy efficiency anomaly root cause diagnosis report. .

7. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The intelligent recommendation of energy efficiency optimization strategies in S6 includes the following steps: S61. Establish a knowledge base for energy efficiency optimization strategies for biomass power plants, including equipment adjustment strategies. Optimization strategies Maintenance and management strategies ; S62. Based on the case-based reasoning technology, recommend appropriate optimization strategies according to the energy efficiency anomaly diagnosis results, and generate a preliminary set of optimization schemes; S63. Employ a multi-objective optimization algorithm to simulate the effectiveness and evaluate the benefits of the recommendation strategy, and determine the optimal energy efficiency improvement path; S64. Prioritize the various optimization strategies using the Analytic Hierarchy Process (AHP) to generate a set of energy efficiency optimization strategies. .

8. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The digital twin technology model construction model in S7 includes the following steps: S71. Construct a digital twin model of a biomass power plant based on physical modeling and data-driven methods. The model includes a virtual-real mapping module and a dynamic simulation module. S72. The energy efficiency optimization scheme is virtually verified using a digital twin model. The verification process includes scheme effect simulation and risk identification operations. S73. Establish a real-time interaction mechanism between energy efficiency monitoring data and digital twin model, wherein the mechanism supports the visualization of energy efficiency status and maintenance early warning; S74. Iterative optimization of energy efficiency optimization strategy using digital twin technology, wherein the optimization process includes strategy parameter adjustment and model version update.

9. The method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The blockchain technology used in S8 to establish a storage mechanism includes the following steps: S81. Construct a distributed storage network for energy efficiency data of biomass power plants based on blockchain. The network consists of multiple nodes, each of which stores a complete copy of the energy efficiency data and synchronizes the data status through a consensus mechanism. S82. Smart contract technology is used to automatically verify and share energy efficiency data. The smart contract includes data format verification rules, access control logic, and data transaction execution conditions. S83. Apply timestamps and hash algorithms to process energy efficiency monitoring data, generate a data hash sequence containing timestamps, and link each data block through a chain structure; S84. Establish a blockchain-based energy efficiency data traceability system, which records the entire chain of energy efficiency data operation logs from collection and storage to access, and supports forward and reverse tracing based on data hash.

10. A method for real-time monitoring and diagnosis of the operating energy efficiency of a biomass power plant according to claim 1, characterized in that: The energy efficiency trend prediction using artificial intelligence algorithms in S9 includes the following steps: S91. Use the LSTM neural network algorithm to predict the trend of energy efficiency of biomass power plants and generate short-term, medium-term and long-term energy efficiency prediction reports. S92. Establish an energy efficiency early warning mechanism based on the prediction results, and set multi-level early warning thresholds and corresponding emergency response plans.

Images