Big Data-Driven Comprehensive Evaluation System and Method for Energy Efficiency

Abstract

The present invention discloses a comprehensive evaluation system and method for energy efficiency driven by big data, which relates to the technical field of energy efficiency evaluation. The parameter determination module estimates the parameters of the Bayesian network model based on the Bayesian inference method. The evaluation module substitutes the factory operation data into the trained Bayesian network model, analyzes the factory operation data through the Bayesian network model, and then evaluates the factory energy efficiency. The management module optimizes and adjusts the Bayesian network model according to the actual situation, formulates an energy management strategy based on the output result of the Bayesian network model, optimizes the production process, and improves the energy utilization efficiency. After comprehensively analyzing the factory operation data based on the Bayesian network model, the evaluation system evaluates the factory energy efficiency and formulates corresponding energy management strategies. The Bayesian network model represents the probabilistic dependence relationship between multiple variables as a network structure, improves the modeling ability for uncertainty, and reduces the evaluation error of the factory energy efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of energy efficiency evaluation, and particularly to a big data-driven comprehensive energy efficiency evaluation system and method. Background Art

[0002] With the development of the global economy and the growth of the population, the energy supply has become increasingly tense. At the same time, energy consumption has also led to serious environmental pollution and climate change problems. Therefore, improving energy utilization efficiency has become an important topic in global energy management today. An evaluation system is a system that uses big data technology and analysis methods to evaluate and improve energy utilization efficiency. At all levels of enterprises and society, energy management has been increasingly emphasized. Through scientific energy management means, reasonable utilization and conservation of resources can be achieved, energy costs can be reduced, the competitiveness of enterprises can be improved, and the impact on the environment can be reduced.

[0003] The existing technologies have the following deficiencies:

[0004] In the big data environment, energy data often comes from multiple different sources and data sets, and there is often a certain degree of uncertainty in energy data, such as measurement errors, data missing, etc. It is very difficult for traditional evaluation systems to effectively handle these uncertainties, resulting in unreliable evaluation results and increased evaluation errors. Summary of the Invention

[0005] The purpose of the present invention is to provide a big data-driven comprehensive energy efficiency evaluation system and method to solve the deficiencies in the background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions: A big data-driven comprehensive energy efficiency evaluation system includes a data acquisition module, a processing module, a parameter determination module, an evaluation module, and a management module;

[0007] Data acquisition module: Collect factory operation data. The factory operation data covers the operation conditions of the factory over a period of time, and the factory operation data is distributed in time and space. The factory operation data is sent to the processing module;

[0008] Processing module: Preprocess the collected data. The preprocessed factory operation data is sent to the evaluation module;

[0009] Parameter determination module: Estimate the parameters of the Bayesian network model through the Bayesian inference method. The parameter information is sent to the evaluation module;

[0010] Evaluation module: Substitute the factory operation data into the trained Bayesian network model, analyze the factory operation data through the Bayesian network model, and then evaluate the factory energy efficiency. The evaluation result is sent to the management module;

[0011] Management module: Optimize and adjust the Bayesian network model according to the actual situation, and formulate an energy management strategy and optimize the production process based on the output results of the Bayesian network model.

[0012] Preferably, the data acquisition module collects factory operation data, and the factory operation data includes production output, equipment operation status, environmental temperature, and energy consumption.

[0013] The processing module preprocesses the collected data, and the preprocessing includes data cleaning, missing value processing, and outlier processing.

[0014] Preferably, the parameter determination module collects factory data over a period of time, including observations of production output, equipment operation status, environmental temperature, and energy consumption. For each node, determine the conditional probability distribution, and use the Gaussian distribution to represent the probability distribution of each node.

[0015] Let the mean of the Gaussian distribution of the production output node be μ1 and the standard deviation be σ1.

[0016] Let the conditional probability distribution of the equipment operation status node depend on the energy consumption. That is, given the energy consumption, the mean of the Gaussian distribution of the equipment operation status is μ2 + β * energy consumption, and the standard deviation is σ2.

[0017] Let the conditional probability distribution of the environmental temperature node depend on the energy consumption. That is, given the energy consumption, the mean of the Gaussian distribution of the environmental temperature is μ3 + γ * energy consumption, and the standard deviation is σ3.

[0018] Let the mean of the Gaussian distribution of the energy consumption node be μ4 and the standard deviation be σ4.

[0019] The mean and standard deviation of the production output node are calculated from the factory operation data, and the parameters of the equipment operation status node, environmental temperature node, and energy consumption node are estimated by fitting, that is, estimating μ2, β, σ2, μ3, γ, σ3, μ4, σ4.

[0020] Preferably, the training of the Bayesian network model includes the following steps:

[0021] Collect the historical operation data of the factory, including production output, equipment operation status, environmental temperature, and energy consumption. Based on the data mining algorithm, construct the structure of the Bayesian network, and determine the dependence relationship between nodes. The functional expression of the Bayesian network model is:

[0022]

[0023] Wherein, P(x) is the production output of node x, P(y|e) is the equipment operation status of node y, P(z|e) is the environmental temperature of node z, P(e) is the energy consumption of node e, N(·) represents the Gaussian distribution, μx is the mean of the production output, and σ x is the standard deviation of the production output, μy is the mean of the equipment operation status under the reference energy consumption, βy is the influence coefficient of the energy consumption on the equipment operation status, and σ y is the standard deviation of the equipment operation status, μz is the mean of the environmental temperature under the reference energy consumption, βz is the influence coefficient of the energy consumption on the environmental temperature, and σ z is the standard deviation of the environmental temperature, μe is the mean of the energy consumption, and σ e is the standard deviation of the energy consumption.

[0024] Verify and evaluate the trained Bayesian network model, check the goodness of fit and prediction performance of the model, and optimize the model according to the results of the model verification and evaluation.

[0025] Preferably, the evaluation module substitutes the operation data of the factory into the trained Bayesian network model, uses the Bayesian network model to calculate the posterior probability distribution of each node based on the given observed data, and evaluates the energy efficiency of the factory by analyzing the relationship between the production output, equipment operation status, environmental temperature, and energy consumption based on the inference results of the Bayesian network model;

[0026] Infer through the Bayesian inference algorithm and calculate the posterior probability distribution of each node based on the given observed data;

[0027] Analyze the relationship between the production output and the energy consumption, determine the energy consumption required for production, analyze the influence of the equipment operation status on the energy consumption, find out whether the equipment operation effectively utilizes the energy, analyze the influence of the environmental temperature on the energy consumption, and identify the influence of environmental factors on the energy efficiency;

[0028] Predict the unobserved variables through the Bayesian network model, including the production output, equipment operation status, environmental temperature, or energy consumption at a future time point, and evaluate the energy efficiency of the factory based on the prediction results.

[0029] Preferably, the management module analyzes the inference results of the Bayesian network model, obtains the posterior probability distribution of each node, and the relationship between the nodes;

[0030] Identify the key factors that have the greatest impact on the energy consumption according to the output results of the Bayesian network model;

[0031] Formulate corresponding energy management strategies according to the identified key influencing factors. The energy management strategies include:

[0032] Arrange the production plan according to the prediction results of production output;

[0033] Formulate the equipment maintenance plan according to the analysis results of the equipment operation status;

[0034] Take corresponding environmental control measures according to the impact of environmental temperature on energy consumption;

[0035] Introduce energy-saving technologies and equipment according to the results of the Bayesian network model.

[0036] Preferably, the data acquisition module collects the factory operation data through data sources, and the factory operation data includes production output, equipment operation status, environmental temperature, and energy consumption;

[0037] The data sources include sensors, monitoring devices, production systems, weather stations, and energy meters inside the factory;

[0038] Based on the preset acquisition frequency, acquisition time period, and acquisition method of the evaluation system, collect the factory operation data in real time, and store the collected data in the database or data warehouse.

[0039] A comprehensive evaluation method for energy efficiency driven by big data, the evaluation method includes the following steps:

[0040] The evaluation system collects the factory operation data, including production output, equipment operation status, environmental temperature, and energy consumption, and preprocesses the collected data;

[0041] Estimate the parameters of the Bayesian network model through the Bayesian inference method, including the probability distribution and conditional probability distribution between nodes;

[0042] Substitute the factory operation data into the trained Bayesian network model, and evaluate the factory energy efficiency after analyzing the factory operation data through the Bayesian network model;

[0043] Optimize and adjust the Bayesian network model according to the actual situation, and formulate the energy management strategy and optimize the production process according to the output results of the model.

[0044] In the above technical solution, the technical effects and advantages provided by the present invention:

[0045] The present invention estimates the parameters of the Bayesian network model by a parameter determination module based on the Bayesian inference method, including the probability distribution, conditional probability distribution, etc. between nodes. The evaluation module substitutes the factory operation data into the trained Bayesian network model, and evaluates the factory energy efficiency after analyzing the factory operation data through the Bayesian network model. The management module optimizes and adjusts the Bayesian network model according to the actual situation, formulates an energy management strategy based on the output result of the Bayesian network model, optimizes the production process, and improves the energy utilization efficiency. After comprehensively analyzing the factory operation data based on the Bayesian network model, the evaluation system evaluates the factory energy efficiency and formulates corresponding energy management strategies. The Bayesian network model represents the probabilistic dependence relationship between multiple variables as a network structure, improves the modeling ability for uncertainty, and reduces the evaluation error of the factory energy efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.

[0047] Figure 1 It is a system module diagram of the present invention.

[0048] Figure 2 It is a method flow chart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0050] Embodiment: Please refer to Figure 1 As shown, the big data-driven comprehensive energy efficiency evaluation system in this embodiment includes a data acquisition module, a processing module, a parameter determination module, an evaluation module, and a management module.

[0051] Data acquisition module: Collect a large amount of factory operation data, including production output, equipment operation status, environmental temperature, energy consumption, etc. The factory operation data covers the factory operation situation within a certain period and should have a certain spatio-temporal distribution. The factory operation data is sent to the processing module.

[0052] Processing module: Preprocess the collected data, including data cleaning, missing value handling, outlier handling, etc. The preprocessed factory operation data is sent to the evaluation module;

[0053] Parameter determination module: Estimate the parameters of the Bayesian network model through Bayesian inference methods, including the probability distribution and conditional probability distribution between nodes. For example, we can use the maximum likelihood estimation method to estimate the conditional probability distribution between nodes, and the parameter information is sent to the evaluation module;

[0054] Evaluation module: Substitute the factory operation data into the trained Bayesian network model, and evaluate the factory energy efficiency after analyzing the factory operation data through the Bayesian network model. The evaluation results are sent to the management module;

[0055] Management module: Optimize and adjust the Bayesian network model according to the actual situation, formulate an energy management strategy based on the output results of the Bayesian network model, optimize the production process, and improve the energy utilization efficiency.

[0056] In this application, the parameter determination module estimates the parameters of the Bayesian network model based on Bayesian inference methods, including the probability distribution and conditional probability distribution between nodes. The evaluation module substitutes the factory operation data into the trained Bayesian network model, and evaluates the factory energy efficiency after analyzing the factory operation data through the Bayesian network model. The management module optimizes and adjusts the Bayesian network model according to the actual situation, formulates an energy management strategy based on the output results of the Bayesian network model, optimizes the production process, and improves the energy utilization efficiency. After comprehensively analyzing the factory operation data based on the Bayesian network model, this evaluation system evaluates the factory energy efficiency and formulates corresponding energy management strategies. The Bayesian network model represents the probabilistic dependence relationship between multiple variables as a network structure, improves the modeling ability for uncertainty, and reduces the evaluation error of the factory energy efficiency.

[0057] Please refer to Figure 2 as shown in

[0058] The evaluation system collects a large amount of factory operation data, including production output, equipment operation status, environmental temperature, energy consumption, etc. The factory operation data covers the factory operation conditions over a certain period and should have a certain spatio-temporal distribution. Preprocess the collected data, including data cleaning, missing value handling, outlier handling, etc. Estimate the parameters of the Bayesian network model through Bayesian inference methods, including the probability distribution and conditional probability distribution between nodes. For example, we can use the maximum likelihood estimation method to estimate the conditional probability distribution between nodes, substitute the factory operation data into the trained Bayesian network model, evaluate the factory energy efficiency after analyzing the factory operation data through the Bayesian network model, and optimize and adjust the Bayesian network model according to the actual situation to improve its accuracy and practicality. Energy management strategies can be formulated based on the output results of the model to optimize the production process and improve energy utilization efficiency.

[0059] Data acquisition module: Collect a large amount of factory operation data, including production output, equipment operation status, environmental temperature, energy consumption, etc. The factory operation data covers the factory operation conditions over a certain period and should have a certain spatio-temporal distribution. The factory operation data is sent to the processing module;

[0060] Data acquisition objective: The objective is to evaluate the energy utilization efficiency of the factory. Therefore, it is necessary to determine what data needs to be collected, including production output, equipment operation status, environmental temperature, energy consumption, etc.

[0061] Select data sources: Determine the sources of data acquisition. The data sources can include sensors, monitoring devices, production systems, etc. inside the factory, and can also include external data sources, such as weather stations, energy meters, etc.

[0062] Design a data acquisition plan: Design a suitable data acquisition plan, including selecting the acquisition frequency, acquisition time period, acquisition method, etc. Determine the specific details of data acquisition according to the actual situation and requirements of the factory.

[0063] Deploy data acquisition equipment: Deploy data acquisition equipment according to the designed plan. This may involve operations such as installing sensors, connecting monitoring devices, and setting up the data acquisition system.

[0064] Real-time data acquisition: Start real-time acquisition of factory operation data. The data acquisition should be continuous and be able to cover the factory operation conditions over a certain period. Ensure the timeliness and accuracy of the data.

[0065] Data storage and management: Store the collected data in a database or data warehouse and perform appropriate management. Ensure the integrity, consistency, and security of the data for convenient subsequent data analysis and processing.

[0066] Data quality monitoring: Regularly monitor data quality, detect and handle issues such as anomalies and missing values in the data. Ensure that the data quality meets the requirements of analysis and modeling.

[0067] Analysis of data spatio-temporal distribution: Analyze the spatio-temporal distribution characteristics of data to understand the distribution patterns of data in time and space. This helps to better understand the operation of the factory and provides a basis for subsequent energy efficiency evaluation.

[0068] Processing module: Preprocess the collected data, including data cleaning, handling missing values, outliers, etc., and send the preprocessed factory operation data to the evaluation module;

[0069] Data cleaning: Data cleaning refers to the initial cleaning of data to remove incomplete, inconsistent, or inaccurate data. This may include removing duplicate data, fixing data format errors, handling incorrect data, etc.;

[0070] Removing duplicate data: First, detect whether there are duplicate records or rows in the data, and if so, delete them. Duplicate data may lead to biased analysis results or waste of computing resources.

[0071] Fixing data format errors: Check whether the data format meets the expectations and fix format errors. For example, incorrect date formats, special characters in text fields, etc. Fixing format errors helps with data uniformity and usability.

[0072] Handling missing values: Handle missing values in the data to ensure data integrity and accuracy. Methods for handling missing values include deleting missing values, filling missing values (such as filling with mean, median, or interpolation);

[0073] Identifying missing values: First, check the data to determine which fields or records have missing values. Missing values can be in the form of blank fields, NaN (Not a Number), NULL, etc.

[0074] Determining the processing strategy: Based on the characteristics of the data and the analysis requirements, determine the specific strategy for handling missing values. Common strategies include deleting missing values, filling missing values, using machine learning models for prediction, etc.

[0075] Deleting missing values: If the number of missing values is small or has little impact on the analysis results, you can choose to delete the records or fields containing missing values. Deleting missing values may reduce the amount of data, but can ensure the accuracy of the analysis results.

[0076] Filling missing values: If deleting missing values will result in too little data or loss of important information, you can choose to fill in the missing values. Common filling methods include:

[0077] Filling with mean or median: For numerical data, the mean or median of the field can be used to fill in missing values. This method is simple and fast, but may cause changes in the data distribution.

[0078] Filling with mode: For categorical data, the mode of the field can be used to fill in missing values. The mode is the value that appears most frequently in the field.

[0079] Interpolation filling: For time series data or data with continuity, interpolation methods such as linear interpolation and polynomial interpolation can be used for filling. This method can better maintain the continuity and trend of the data.

[0080] Outlier handling: Detect and handle outliers in the data to prevent them from having an adverse impact on the analysis results. Methods for outlier handling include deleting outliers, repairing outliers (such as smoothing, replacing with appropriate values), etc.;

[0081] Identifying outliers: First, the data needs to be examined to identify potential outliers. Outliers may appear as values that are significantly different from other data points or values that are far from the central tendency of the data set.

[0082] Determining outlier criteria: Based on the characteristics of the data and domain knowledge, determine the criteria or thresholds for outliers. This can be based on statistical methods such as mean plus or minus a certain number of standard deviations, or on professional knowledge and experience.

[0083] Detecting outliers: Use the determined outlier criteria to detect the data and find the data points that meet the outlier conditions. This can be achieved through programming methods or through visual inspection using visualization tools.

[0084] Handling outliers: Once outliers are detected, one of the following methods can be chosen to handle them:

[0085] Deleting outliers: If outliers have a significant negative impact on the analysis results and cannot be explained or corrected, considering directly deleting the outliers.

[0086] Repairing outliers: Repair the outliers to make them conform to the data distribution pattern. Repair methods can include smoothing (such as using moving average or median smoothing), replacing with appropriate values (such as using interpolation methods or estimating based on adjacent data points), converting to values within an acceptable range, etc.

[0087] Data transformation: Transform the data to meet the requirements of subsequent analysis and modeling. This may include operations such as standardizing, normalizing, and logarithmic transformation of the data to make the data more conform to the assumptions of analysis and modeling.

[0088] Data integration: Integrate data from different data sources so that the data can be analyzed and processed in the same dataset. This may involve operations such as unifying data fields and converting data formats.

[0089] Data dimensionality reduction: Perform dimensionality reduction operations on the data to reduce data complexity and redundancy. This includes feature selection to improve the efficiency and accuracy of subsequent analysis and modeling.

[0090] Parameter determination module: Estimate the parameters of the Bayesian network model through Bayesian inference methods, including probability distributions and conditional probability distributions between nodes. For example, we can use the maximum likelihood estimation method to estimate the conditional probability distribution between nodes, and send the parameter information to the evaluation module;

[0091] The influencing factors of factory energy efficiency include production output (Production), equipment operating status (Equipment-Status), environmental temperature (Temperature), and energy consumption (Energy-Consumption);

[0092] The Bayesian network model has four nodes, namely production output (Production), equipment operating status (Equipment-Status), environmental temperature (Temperature), and energy consumption (Energy-Consumption). In the Bayesian network, it is assumed that production output is affected by equipment operating status, environmental temperature, and energy consumption, and equipment operating status and environmental temperature are also affected by energy consumption;

[0093] Collect factory data over a period of time, including observed values of production output, equipment operating status, environmental temperature, and energy consumption. For each node, we need to determine its conditional probability distribution. We can use the Gaussian distribution to represent the probability distribution of each node. Specifically:

[0094] For the production output node (Production), assume that the mean of its Gaussian distribution is μ1 and the standard deviation is σ1.

[0095] For the equipment operating status node (Equipment-Status), assume that its conditional probability distribution depends on energy consumption (Energy-Consumption). That is, given energy consumption, the mean of the Gaussian distribution of equipment operating status is μ2 + β * energy consumption, and the standard deviation is σ2.

[0096] For the environmental temperature node (Temperature), assume that its conditional probability distribution depends on the energy consumption (Energy-Consumption), i.e., given the energy consumption, the mean of the Gaussian distribution of the environmental temperature is μ3 + γ * energy consumption, and the standard deviation is σ3.

[0097] For the energy consumption node (Energy-Consumption), assume that its Gaussian distribution has a mean of μ4 and a standard deviation of σ4;

[0098] Estimate the parameters of each node. Specifically:

[0099] The mean and standard deviation of the production output node (Production) can be directly calculated from the data.

[0100] The parameters of the equipment operation status node (Equipment-Status), environmental temperature node (Temperature), and energy consumption node (Energy-Consumption) need to be estimated by fitting, i.e., estimate μ2, β, σ2, μ3, γ, σ3, μ4, σ4.

[0101] Parameter estimation of the equipment operation status node (Equipment-Status):

[0102] The conditional probability distribution of the equipment operation status node depends on the energy consumption. We assume that the equipment operation status follows a Gaussian distribution, whose mean is affected by the energy consumption, expressed as μ2 + β * energy consumption, and the standard deviation is σ2. Therefore, we need to estimate the parameters μ2, β, σ2.

[0103] μ2: The average value of the equipment operation status at a certain reference energy consumption.

[0104] β: The influence coefficient of energy consumption on the equipment operation status.

[0105] σ2: The standard deviation of the equipment operation status.

[0106] Parameter estimation can use the least squares method or other regression methods to fit the relationship between energy consumption and equipment operation status.

[0107] Parameter estimation of the environmental temperature node (Temperature):

[0108] The conditional probability distribution of the environmental temperature node also depends on the energy consumption. We also assume that the environmental temperature follows a Gaussian distribution, whose mean is affected by the energy consumption, expressed as μ3 + γ * energy consumption, and the standard deviation is σ3. Therefore, we need to estimate the parameters μ3, γ, σ3.

[0109] μ3: The average value of the environmental temperature at a certain reference energy consumption.

[0110] γ: The influence coefficient of energy consumption on environmental temperature.

[0111] σ3: The standard deviation of the environmental temperature.

[0112] Similarly, parameter estimation can use the least squares method or other regression methods to fit the relationship between energy consumption and environmental temperature.

[0113] Parameter estimation of the Energy-Consumption node:

[0114] The probability distribution of the Energy-Consumption node itself needs to estimate its mean μ4 and standard deviation σ4.

[0115] μ4: The average value of energy consumption.

[0116] σ4: The standard deviation of energy consumption.

[0117] Evaluation module: Substitute the factory operation data into the trained Bayesian network model, evaluate the factory energy efficiency after analyzing the factory operation data through the Bayesian network model, and send the evaluation results to the management module;

[0118] Substitute the factory operation data into the trained Bayesian network model and use the Bayesian network model for inference, that is, calculate the posterior probability distribution of each node based on the given observed data. This includes predicting unobserved variables. Based on the model inference results, evaluate the factory energy efficiency by analyzing the relationship between production output, equipment operation status, environmental temperature, and energy consumption;

[0119] Inference: Use the Bayesian network model for inference, that is, calculate the posterior probability distribution of each node based on the given observed data. This can be achieved through Bayesian inference algorithms such as variable elimination, sampling methods (such as Markov chain Monte Carlo method), etc.

[0120] Energy efficiency evaluation: Based on the model inference results, evaluate the factory energy efficiency by analyzing the relationship between production output, equipment operation status, environmental temperature, and energy consumption. The following aspects can be considered:

[0121] Relationship between production output and energy consumption: Analyze the relationship between production output and energy consumption to determine the energy consumption required for production.

[0122] Relationship between equipment operation status and energy consumption: Analyze the impact of equipment operation status on energy consumption to find out whether the equipment operation effectively utilizes energy.

[0123] Impact of ambient temperature on energy consumption: Analyze the impact of ambient temperature on energy consumption and identify the influence of environmental factors on energy efficiency.

[0124] Prediction of unobserved variables: Use a Bayesian network model to predict unobserved variables, such as production output, equipment operating status, ambient temperature, or energy consumption at a future point in time.

[0125] Equipment with excessive energy consumption:

[0126] Identify which equipment has a high energy consumption, which may be due to equipment aging, lack of timely maintenance, unreasonable process flow, etc.

[0127] Analyze the energy consumption of these equipment, find out the specific reasons for excessive energy consumption, and formulate corresponding improvement measures.

[0128] Energy waste caused by environmental factors:

[0129] Analyze the impact of environmental factors (such as temperature, humidity, etc.) on energy consumption and identify whether there is energy waste caused by environmental changes.

[0130] For equipment or process steps that are greatly affected by the environment, consider taking measures such as adjusting process parameters, improving equipment design, and strengthening environmental control to reduce energy waste.

[0131] Energy waste caused by poor equipment operating status:

[0132] Analyze the impact of the equipment operating status on energy consumption and identify whether there are problems such as unstable equipment operation and low efficiency.

[0133] For equipment with poor operating status, it is necessary to diagnose and find out the reasons, and may need to take measures such as equipment maintenance, replacement of key components, and optimization of equipment scheduling to improve the energy utilization efficiency of the equipment.

[0134] Energy waste caused by unreasonable process flow:

[0135] Analyze the energy consumption of the process flow and identify whether there are problems such as unreasonable process parameter settings and energy losses during the production process.

[0136] For process steps with problems, consider optimizing the process flow, improving production technology, and introducing energy-saving technologies to reduce energy consumption and improve energy utilization efficiency.

[0137] The training of the Bayesian network model includes the following steps:

[0138] Data collection: First, collect the historical operation data of the factory, including indicators such as production output, equipment operation status, environmental temperature, and energy consumption. These data should cover the operation of the factory over a period of time and should have a certain spatio-temporal distribution.

[0139] Construct the Bayesian network structure: Based on data mining algorithms, construct the structure of the Bayesian network, determine the dependencies between nodes. The functional expression of the Bayesian network model is:

[0140]

[0141] In the formula, P(x) is the production output of node x, P(y|e) is the equipment operation status of node y, P(z|e) is the environmental temperature of node z, P(e) is the energy consumption of node e, N(·) represents the Gaussian distribution, μx is the mean of the production output, σ x is the standard deviation of the production output, μy is the mean of the equipment operation status under the reference energy consumption, βy is the influence coefficient of energy consumption on the equipment operation status, σ y is the standard deviation of the equipment operation status, μz is the mean of the environmental temperature under the reference energy consumption, βz is the influence coefficient of energy consumption on the environmental temperature, σ z is the standard deviation of the environmental temperature, μe is the mean of the energy consumption, σ e is the standard deviation of the energy consumption.

[0142] Model validation and evaluation: Validate and evaluate the trained Bayesian network model, and check the goodness of fit and prediction performance of the model. Methods such as cross-validation and information criteria can be used to evaluate the performance of the model.

[0143] Model tuning: According to the results of model validation and evaluation, tune the model, optimize the structure and parameters of the model, and improve the performance and generalization ability of the model.

[0144] Management module: Optimize and adjust the Bayesian network model according to the actual situation, formulate an energy management strategy based on the output results of the Bayesian network model, optimize the production process, and improve energy utilization efficiency;

[0145] Analyze the output results of the Bayesian network model: First, conduct a detailed analysis of the inference results of the Bayesian network model to understand the posterior probability distribution of each node and the relationships between nodes.

[0146] Identify key influencing factors: According to the output results of the Bayesian network model, identify the key factors that have the greatest impact on energy consumption. These may include factors such as production output, equipment operation status, and environmental temperature.

[0147] Develop an energy management strategy: Based on the identified key influencing factors, develop corresponding energy management strategies. This may include the following aspects:

[0148] Optimize the production plan: According to the predicted production output, reasonably arrange the production plan to avoid overcapacity or undercapacity, so as to reduce energy waste.

[0149] Equipment maintenance and update: Based on the analysis results of the equipment operation status, formulate an equipment maintenance plan, regularly inspect and maintain the equipment to ensure stable equipment operation and reduce energy consumption.

[0150] Environmental control measures: According to the impact of environmental temperature on energy consumption, take corresponding environmental control measures, such as adjusting the air conditioning system, improving insulation materials, etc., to reduce energy consumption.

[0151] Application of energy-saving technologies: According to the results of the Bayesian network model, introduce energy-saving technologies and equipment, such as energy efficiency transformation, update of energy-saving equipment, etc., to improve energy utilization efficiency.

[0152] Suppose we have a factory producing auto parts and want to optimize energy utilization efficiency. We use a Bayesian network model to analyze the operation data of the factory and develop an energy management strategy based on the model output results, including the following steps:

[0153] Analyze the output results of the Bayesian network model:

[0154] After training and inference, the Bayesian network model gives the posterior probability distribution of each node, including the relationship between production output, equipment operation status, environmental temperature, and energy consumption.

[0155] Identify key influencing factors:

[0156] Analyze the model output results to determine which factors have the greatest impact on energy consumption, such as production output, equipment operation status, etc.

[0157] Develop an energy management strategy:

[0158] Optimize the production plan: According to the predicted production output and market demand, reasonably arrange the production plan to avoid overcapacity or undercapacity, so as to reduce energy waste.

[0159] Equipment maintenance and update: Based on the analysis results of the equipment operation status, formulate an equipment maintenance plan, regularly inspect and maintain the equipment to ensure stable equipment operation and reduce energy consumption.

[0160] Environmental control measures: According to the impact of environmental temperature on energy consumption, adjust the air conditioning system, improve insulation materials, etc., to reduce energy consumption.

[0161] Application of energy-saving technologies: Introduce energy-saving technologies and equipment, such as LED lighting, high-efficiency equipment, intelligent control systems, etc., to improve energy utilization efficiency.

[0162] Implement energy management strategies:

[0163] Implement the formulated energy management strategies into the actual production process to ensure the effective execution of the strategies. Implement the equipment maintenance plan, adjust the production plan and process parameters, update the equipment and introduce energy-saving technologies.

[0164] Continuous monitoring and optimization:

[0165] Continuously monitor and evaluate the implemented energy management strategies, promptly identify problems and make adjustments and optimizations. Regularly check the energy consumption situation, production data and equipment status, and adjust the management strategies according to the monitoring results to further improve energy utilization efficiency.

[0166] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulations to obtain a formula that is closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0167] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0168] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not elaborate on all the details, nor do they limit the present invention to only the specific implementation manners. Obviously, many modifications and variations can be made according to the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the present invention, so that those skilled in the art in the relevant technical field can well understand and utilize the present invention. The present invention is only limited by the claims and their full scope and equivalents.

Claims

1. A comprehensive evaluation system for energy efficiency driven by big data, characterized in that: It includes a data acquisition module, a processing module, a parameter determination module, an evaluation module, and a management module; Data acquisition module: It is used to collect the factory operation data, and the factory operation data is sent to the processing module; Processing module: It is used to preprocess the collected data, and the preprocessed factory operation data is sent to the evaluation module; Parameter determination module: It is used to estimate the parameters of the Bayesian network model by the Bayesian inference method, and the parameter information is sent to the evaluation module; Evaluation module: It is used to substitute the factory operation data into the trained Bayesian network model, evaluate the factory energy efficiency after analyzing the factory operation data through the Bayesian network model, and send the evaluation result to the management module; Management module: It is used to optimize and adjust the Bayesian network model according to the actual situation, and formulate an energy management strategy and optimize the production process according to the output result of the Bayesian network model; In the said parameter determination module, The Bayesian network model has four nodes, namely production output, equipment operation status, environmental temperature, and energy consumption; Collect the observed values of production output, equipment operation status, environmental temperature, and energy consumption. For each node, determine the conditional probability distribution, and use the Gaussian distribution to represent the probability distribution of each node; Let the mean of the Gaussian distribution of the production output node be μ1, and the standard deviation be σ1; Let the conditional probability distribution of the equipment operation status node depend on the energy consumption, that is, under the condition of a given energy consumption, the mean of the Gaussian distribution of the equipment operation status is μ2 + β * energy consumption, and the standard deviation is σ2; Let the conditional probability distribution of the environmental temperature node depend on the energy consumption, that is, under the condition of a given energy consumption, the mean of the Gaussian distribution of the environmental temperature is μ3 + γ * energy consumption, and the standard deviation is σ3; Let the mean of the Gaussian distribution of the energy consumption node be μ4, and the standard deviation be σ4; The mean and standard deviation of the production output node are calculated from the factory operation data, and the parameters of the equipment operation status node, environmental temperature node, and energy consumption node are estimated by fitting; The training of the said Bayesian network model includes the following steps: Collect the historical operation data of the factory, including production output, equipment operation status, environmental temperature, and energy consumption. Based on the data mining algorithm, construct the structure of the Bayesian network, determine the dependence relationship between nodes, and the function expression of the Bayesian network model is: Where P(x) is the production output of node x, P(y|e) is the equipment operating status of node y, P(z|e) is the environmental temperature of node z, P(e) is the energy consumption of node e, N(·) represents the Gaussian distribution, μx is the mean of the production output, and σ x is the standard deviation of the production output, μy is the mean of the equipment operating status under the reference energy consumption, βy is the influence coefficient of the energy consumption on the equipment operating status, and σ y is the standard deviation of the equipment operating status, μz is the mean of the environmental temperature under the reference energy consumption, z is the influence coefficient of the energy consumption on the environmental temperature, and σ z is the standard deviation of the environmental temperature, μe is the mean of the energy consumption, and σ e is the standard deviation of the energy consumption; The evaluation module substitutes the operation data of the factory into the trained Bayesian network model, uses the Bayesian network model to calculate the posterior probability distribution of each node based on the given observed data, and evaluates the energy efficiency of the factory by analyzing the relationship between production output, equipment operation status, environmental temperature, and energy consumption based on the inference result of the Bayesian network model; Perform inference through the Bayesian inference algorithm, and calculate the posterior probability distribution of each node based on the given observed data; Analyze the relationship between production output and energy consumption, determine the energy consumption required for production, analyze the impact of equipment operation status on energy consumption, find out whether the equipment operation effectively utilizes energy, analyze the impact of environmental temperature on energy consumption, and identify the impact of environmental factors on energy efficiency; Predict unobserved variables through a Bayesian network model, including production output, equipment operating status, environmental temperature, or energy consumption at future time points, and evaluate the energy efficiency of the factory based on the prediction results; The management module analyzes the inference results of the Bayesian network model to obtain the posterior probability distribution of each node and the relationships between nodes; Identify the key factors that have the greatest impact on energy consumption based on the output results of the Bayesian network model; Formulate corresponding energy management strategies according to the identified key influencing factors. The energy management strategies include: Arrange the production plan according to the prediction results of production output; Formulate an equipment maintenance plan according to the analysis results of equipment operating status; Take corresponding environmental control measures according to the impact of environmental temperature on energy consumption; Introduce energy-saving technologies and equipment according to the results of the Bayesian network model.

2. The big data-driven comprehensive energy efficiency evaluation system according to claim 1, wherein: The factory operation data includes production output, equipment operating status, environmental temperature, and energy consumption; The processing module preprocesses the collected data, and the preprocessing includes data cleaning, missing value processing, and outlier processing.

3. The comprehensive evaluation system for energy efficiency driven by big data according to claim 1, wherein: The data acquisition module collects factory operation data through data sources, The data sources include sensors, monitoring devices, production systems, weather stations, and energy meters inside the factory; Based on the preset collection frequency, collection time period, and collection method of the evaluation system, collect factory operation data in real time and store the collected data in a database or data warehouse.

4. The comprehensive evaluation method for energy efficiency driven by big data is implemented through the evaluation system described in any one of claims 1-3, and is characterized in that: The evaluation method includes the following steps: The evaluation system collects factory operation data, including production output, equipment operating status, environmental temperature, and energy consumption, and preprocesses the collected data; Estimate the parameters of the Bayesian network model through the Bayesian inference method, including the probability distribution and conditional probability distribution between nodes; Substitute the factory operation data into the trained Bayesian network model, and evaluate the energy efficiency of the factory after analyzing the factory operation data through the Bayesian network model; Optimize and adjust the Bayesian network model according to the actual situation, and formulate energy management strategies and optimize the production process according to the output results of the model; The Bayesian network model has four nodes, namely production output, equipment operating status, environmental temperature, and energy consumption; Collect the observed values of production output, equipment operating status, environmental temperature, and energy consumption. For each node, determine the conditional probability distribution, and use the Gaussian distribution to represent the probability distribution of each node; Let the mean of the Gaussian distribution of the production output node be μ1 and the standard deviation be σ1; Let the conditional probability distribution of the equipment operating status node depend on the energy consumption. That is, given the energy consumption, the mean of the Gaussian distribution of the equipment operating status is μ2 + β * energy consumption, and the standard deviation is σ2; Let the conditional probability distribution of the environmental temperature node depend on the energy consumption. That is, given the energy consumption, the mean of the Gaussian distribution of the environmental temperature is μ3 + γ * energy consumption, and the standard deviation is σ3; Let the mean of the Gaussian distribution of the energy consumption node be μ4 and the standard deviation be σ4; The mean and standard deviation of the production output node are calculated from the factory operation data, and the parameters of the equipment operation status node, the environmental temperature node, and the energy consumption node are estimated by fitting. The training of the Bayesian network model includes the following steps: Collect the historical operation data of the factory, including production output, equipment operation status, environmental temperature, and energy consumption. Based on the data mining algorithm, construct the structure of the Bayesian network, determine the dependencies between nodes. The functional expression of the Bayesian network model is: Where P(x) is the production output of node x, P(y|e) is the equipment operating status of node y, P(z|e) is the environmental temperature of node z, P(e) is the energy consumption of node e, N(·) represents the Gaussian distribution, μx is the mean of the production output, and σ x is the standard deviation of the production output, μy is the mean of the equipment operating status under the reference energy consumption, βy is the influence coefficient of the energy consumption on the equipment operating status, and σ y is the standard deviation of the equipment operating status, μz is the mean of the environmental temperature under the reference energy consumption, z is the influence coefficient of the energy consumption on the environmental temperature, and σ z is the standard deviation of the environmental temperature, μe is the mean of the energy consumption, and σ e is the standard deviation of the energy consumption; Substitute the operation data of the factory into the trained Bayesian network model. Use the Bayesian network model to calculate the posterior probability distribution of each node based on the given observed data. Based on the inference results of the Bayesian network model, evaluate the energy efficiency of the factory by analyzing the relationships among production output, equipment operation status, environmental temperature, and energy consumption. Perform inference through the Bayesian inference algorithm to calculate the posterior probability distribution of each node based on the given observed data. Analyze the relationship between production output and energy consumption to determine the energy consumption required for production. Analyze the impact of equipment operation status on energy consumption to find out whether the equipment operation effectively utilizes energy. Analyze the impact of environmental temperature on energy consumption to identify the impact of environmental factors on energy efficiency. Use the Bayesian network model to predict unobserved variables, including production output, equipment operation status, environmental temperature, or energy consumption at future time points. Evaluate the energy efficiency of the factory based on the prediction results. Analyze the inference results of the Bayesian network model to obtain the posterior probability distribution of each node and the relationships between nodes. Based on the output results of the Bayesian network model, identify the key factors that have the greatest impact on energy consumption. According to the identified key influencing factors, formulate corresponding energy management strategies. The energy management strategies include: Arrange the production plan according to the prediction results of production output. Formulate the equipment maintenance plan according to the analysis results of equipment operation status. Take corresponding environmental control measures according to the impact of environmental temperature on energy consumption. Introduce energy-saving technologies and equipment according to the results of the Bayesian network model.

Images