Heat supply system digital production management platform based on artificial intelligence

Abstract

The invention discloses a heat supply system digital production management platform based on artificial intelligence, and the platform comprises an inspection management module, an overhaul and maintenance management module, and an emergency safety management module which are respectively used for formulating an inspection work order, an overhaul and maintenance work order, and an emergency guide work order; the multi-source work order integrated management module is used for calling a work order identification large model to obtain a work order content identification result after obtaining multi-source work order integrated information, establishing a work order distribution optimization model in combination with a work order portrait, a work order processing personnel portrait and a current personnel state, and obtaining optimal personnel information for executing each work order; the production management mobile equipment is used for personnel to receive and execute various types of distributed work orders and monitor track information and work order processing states of the whole execution process of the various types of work orders; and the production evaluation management module is used for setting comprehensive evaluation multi-type indexes, carrying out comprehensive evaluation on the production management condition in the heating season, analyzing an optimizable mechanism in the work order processing process and formulating a work order processing optimization strategy.

Description

Technical Field

[0001] This invention belongs to the field of heating system technology, specifically relating to a digital production management platform for heating systems based on artificial intelligence. Background Technology

[0002] As a core component of urban infrastructure, the heating system directly affects residents' quality of life and the stability of urban operations. Its production and management level not only influences the quality of heating services but is also closely related to energy consumption and safe operation. With the acceleration of urbanization and the continuous expansion of heating coverage, the heating network is becoming increasingly complex. The traditional management model, which relies on manual experience, is no longer sufficient to meet the needs of large-scale and refined management.

[0003] Heating system production management involves multiple functional modules, including inspection management, maintenance management, emergency safety management, and work order creation and dispatch. However, current traditional inspection and maintenance rely on manual experience, lack dynamic response and standardized testing methods, leading to over-maintenance or untimely repairs. Furthermore, after an accident, accident location and severity assessment are time-consuming, and emergency resource allocation is often inadequate, potentially expanding the scope of the accident and extending the duration of heating outages. In addition, the operation of heating systems generates various types of work orders, including inspection, maintenance, and emergency orders. How to integrate and unify the management of these work orders, and how to dispatch them based on multiple factors such as work order content, personnel skills, response time, and workload to improve work order processing efficiency are urgent problems to be solved.

[0004] Based on the aforementioned technical issues, a new artificial intelligence-based digital production management platform for heating systems needs to be designed. Summary of the Invention

[0005] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and provide a digital production management platform for heating systems based on artificial intelligence. This platform can accurately generate inspection work orders, maintenance work orders, and emergency guidance work orders, manage multiple types of work orders in a unified manner, optimize work order dispatch by using work order profiles and work order processing personnel profiles, comprehensively evaluate the production management situation during the heating season, analyze the mechanisms that can be optimized in the work order processing process, and formulate work order processing optimization strategies.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: This invention provides an artificial intelligence-based digital production management platform for heating systems, comprising: The inspection management module is used to obtain relevant information about the heating system inspection, call the pre-set heating system anomaly prediction model, and formulate inspection work orders in combination with sudden emergency events in the production process. The inspection and maintenance management module is used to call the pre-set heating equipment defect detection model to determine whether the heating equipment has physical defects, and to analyze the severity of the physical defects and the scope of their impact on heating operation based on the detection results, and to formulate inspection and maintenance work orders. The emergency safety management module is used to acquire multimodal data from the accident site when an accident occurs in the heating system, analyze the severity of the accident, the location of the accident, the damage to the equipment, and formulate emergency guidance work orders in combination with social public opinion factors; The multi-source work order integration management module connects with the inspection management module, maintenance management module, and emergency safety management module. It is used to acquire multi-source work order integration information, including inspection work orders, maintenance work orders, and emergency guidance work orders. After that, it calls the pre-trained work order recognition model to obtain the work order content recognition result and sends it to the corresponding work order management personnel for review. Combining the pre-built work order profile, work order processing personnel profile, and current personnel status, it establishes a work order dispatch optimization model to obtain the optimal personnel information for executing each type of work order. The production management mobile device connects to the multi-source work order integration management module to receive and execute various types of work orders dispatched by personnel, monitor the trajectory information and work order processing status of the entire execution process of each type of work order, and provide feedback to the multi-source work order integration management module. The production evaluation management module connects with the multi-source work order integration management module to obtain the processing status of various types of work orders in the current heating season, set comprehensive evaluation indicators for multiple types, conduct a comprehensive evaluation of the production management situation in the heating season, analyze the mechanisms that can be optimized in the work order processing process, and formulate work order processing optimization strategies.

[0007] Furthermore, the process of acquiring relevant information about the heating system inspection and calling a pre-set heating system anomaly prediction model, combined with unexpected emergencies during the production process, to formulate inspection work orders includes: Obtain relevant information for heating system inspections, including physical information of heating equipment, operating data, and historical inspection records; Acquire information on sudden emergencies during the production process, including cold wave weather warnings and equipment malfunction alarms; Based on information related to heating system inspections and sudden emergencies in the production process, machine learning algorithms are used to extract abnormal features and train models to establish a heating system anomaly prediction model. A pre-set heating system anomaly prediction model is invoked to predict whether the heating system will experience potential anomalies. If no potential anomalies are found, a routine inspection work order is created. Otherwise, in addition to creating routine inspection work orders, a special inspection work order is created based on the type, location, and anomaly data of the potential anomalies. The inspection work order includes the inspection cycle, inspection route, and inspection tasks. The types of potential anomalies include abnormal equipment parameters and functional abnormalities during system operation, without causing substantial damage.

[0008] Furthermore, the process involves calling a pre-set heating equipment defect detection model to determine whether the heating equipment has physical defects, and analyzing the severity of the physical defects and their impact on heating operation based on the detection results, and then creating a maintenance work order, including: By calling the multi-source data acquisition submodule to collect basic information, operating data, historical defects and maintenance records of heating equipment, and by traversing and locating relevant technical information on heating maintenance and defect detection from network channels, a heating equipment defect database is constructed. Set up a prompt template for physical defect detection of heating equipment: clearly define the defect type, detection task, detection characteristics, and output defect detection result format; The database of heating equipment defects is input into a pre-trained large model. The model is then guided by prompt word templates to identify physical defects of heating equipment and extract defect feature classifications through semantic understanding, visual analysis and feature matching. The output of defect detection results includes whether physical defects have occurred, the location of the defect in the equipment, the description features of the defect type, the cause features of the defect and the range of the defect's impact. The large language model is used to perform semantic similarity analysis on the output defect features, and redundant features are clustered and merged. The defect detection results containing defect features are transformed into core input variables for defect risk modeling, and a defect risk model is established. Based on the defect risk model, risk values ​​are output, and risk levels are classified according to the risk values. When a defect is classified as high risk, its maintenance is defined as high priority and personnel are immediately assigned to handle it. When a defect is classified as medium or low risk, maintenance is scheduled during off-peak heating periods in conjunction with the heating production scheduling plan. Repair and maintenance work orders are formulated based on the equipment location of the defect, the characteristics of the defect type, the characteristics of the cause of the defect, the characteristics of the scope of the defect's impact, and the risk level.

[0009] Furthermore, the basic information of the heating equipment shall at least include the model, installation time, relevant parameters, and maintenance records of the heat source unit, heat exchanger, heating network pipeline, pumps and valves; the operating data of the heating equipment shall at least include the equipment's start-up and shutdown status, load changes, heat metering parameters, and vibration parameters; the historical defects and maintenance records of the heating equipment shall at least include the type of physical defect, the cause of the defect, the influencing factors of the defect, the handling method, and the repair effect. The types of physical defects in the equipment include at least the following: burner nozzle blockage in the heat source unit; corrosion or leakage in the fuel supply pipeline; wear and abnormal noise in auxiliary equipment; perforation of the heat exchanger tube wall; loose parts; scaling of the heat exchanger tube; blockage of the heat exchange medium channel; seal failure; corrosion of the inner and outer walls of the heating network pipeline; mechanical damage to the pipeline; detachment or damage of the insulation layer; pump body wear; bearing damage; internal leakage of valves; valve jamming; valve body cracks or weld leakage.

[0010] Furthermore, the defect risk model is expressed as: ; This represents the risk value for the i-th type of heating defect. This represents the probability of a defect occurring. The severity of the loss caused by the defect; Let be the probability density function of the probability of defect occurrence; Let be the probability density function of the severity of the loss.

[0011] Furthermore, the maintenance management module is also used to predict the demand for maintenance materials based on historical maintenance data, equipment operating status, and maintenance work orders, using a combination of time series analysis algorithm and association rule mining algorithm, to optimize material inventory management and procure and reserve spare parts resources.

[0012] Furthermore, when an accident occurs in the heating system, multimodal data from the accident site is acquired to analyze the severity of the accident, its location, and the extent of equipment damage. Combined with public opinion factors, emergency guidance work orders are formulated, including: When a sudden, destructive accident occurs in the heating system due to equipment failure, operational error, or external factors, relevant data caused by the accident will be obtained, including historical accident data, real-time abnormal system data, images of equipment damage caused by the accident, geographic information, and heating accident reports from social media and news platforms. After retrieving relevant accident entities from the acquired data and standards, emergency plans, and risk prevention manuals based on the heating system using a multimodal large model, the causal relationship between the accident entities, the temporal relationship between the accident and time, the locational relationship between the accident and geography, and the relationship between the accident and emergency repair measures are determined. Then, a knowledge graph of heating accidents is constructed with accident entities as nodes and relationships as edges. Based on the knowledge graph of heating accidents, accident chain nodes are identified, including the cause node that triggers the accident, the accident occurrence node, the diffusion node that triggers secondary accidents, and the accident response node. At the same time, the impact of the accident on the heating area, the duration of the heating outage, and the public opinion intensity are analyzed, and emergency guidance work orders are formulated, including emergency repair equipment, emergency repair location, personnel allocation, and public opinion response.

[0013] Furthermore, after acquiring multi-source work order integrated information including inspection work orders, maintenance work orders, and emergency guidance work orders, a pre-trained work order recognition model is invoked to obtain work order content recognition results, which are then sent to the corresponding work order management personnel for review. Combining pre-built work order profiles, work order processing personnel profiles, and the current personnel status, a work order dispatch optimization model is established to obtain the optimal personnel information for executing each type of work order, including: Acquire integrated work order information from multiple sources, including inspection work orders, maintenance work orders, and emergency guidance work orders, with at least one work order of each type; After pre-training and fine-tuning the work order recognition model using historical multi-source work order integrated information, the acquired multi-source work order integrated information is input into the work order recognition model. The work order content recognition results, including work order type tags, heating area tags involved in the work order, core business content tags of the work order, and work order priority tags, are identified through semantic understanding and then sent to the corresponding work order management personnel for manual review. After manual review and approval, the work order dispatch submodule is entered to obtain the responsiveness requirements, personnel skill and qualification requirements, and special scenario requirements of each work order as work order service preference features. The content recognition results of each work order are used as the basic features of the work order. Based on the basic features and service preference features of the work order, a profile of each work order is constructed. The basic characteristics and service capability characteristics of each work order processing personnel are obtained to construct a profile of each work order processing personnel. The basic characteristics of the work order processing personnel include skill tags, qualification level tags, current work status, and historical work order processing records. The service capability characteristics of the work order processing personnel include the completion rate of the personnel's work orders, the response speed of the personnel's work orders, the matching degree of the personnel's skill certification, and the personnel's experience in handling special scenarios. Set up feature mapping matching rules between work order profiles and work order processing personnel profiles, perform preliminary quantitative matching between work orders and personnel to form a candidate pool of personnel for each work order processing, and analyze the personnel matching degree by combining the calculated feature weights and feature matching scores of each dimension. Through optimization algorithm, select the personnel with the highest matching degree and balanced work order load rate from the candidate pool as the optimal personnel for work order processing.

[0014] Furthermore, the process involves obtaining the processing status of various types of work orders during the current heating season, setting comprehensive evaluation indicators for multiple types, conducting a comprehensive evaluation of the production management situation during the heating season, analyzing the mechanisms that can be optimized in the work order processing process, and formulating work order processing optimization strategies, including: Acquire full-process data on the creation, response, processing, and feedback of various types of work orders during the current heating season, and set up a comprehensive evaluation of multiple indicators, including work order creation compliance, work order processing efficiency, processing quality, resource utilization efficiency, and risk prevention and control capabilities. The weights of each type of indicator are determined by the analytic hierarchy process (AHP). The scores of each type of indicator are combined and weighted to obtain the total comprehensive evaluation score. Based on the total comprehensive evaluation score, the production management situation during the heating season is classified into different levels. At the same time, a production management evaluation report for the heating season is generated, including the comprehensive evaluation score, the scores of each indicator, and the labeling of the weakest indicators. Based on the identified shortcomings, a root cause analysis was conducted to identify the key factors affecting the indicator scores, including non-compliant work order content and unreasonable personnel assignment, and corresponding work order processing optimization strategies were developed.

[0015] Furthermore, the compliance indicators for work order formulation include work order element completeness rate, work order classification accuracy rate, and work order priority matching rate; the work order processing efficiency indicators include average response time, response timeliness rate, average processing time, and overdue completion rate; the processing quality indicators include acceptance pass rate, defect / accident recurrence rate, and work order acceptance score; the resource utilization efficiency indicators include personnel load balance, average work order processing volume per person, and material consumption deviation rate; and the risk prevention and control capability indicators include defect identification accuracy rate, defect-accident conversion rate, emergency response success rate, and major accident occurrence rate.

[0016] The beneficial effects of this invention are: (1) Based on the abnormal prediction model and the linkage of sudden emergency events, this invention realizes the on-demand formulation of inspection work orders, improves the pertinence and efficiency of inspections, and captures abnormal signals of equipment in advance. Through preventive inspections, it reduces the formation of defects and reduces the failure rate from the source. (2) The defect detection model of the present invention identifies physical defects of equipment and assesses their severity and scope of impact, formulates maintenance work orders, avoids over-maintenance or under-maintenance, rationally allocates resources, improves defect repair efficiency, and reduces the impact on heating quality. (3) The emergency safety module of this invention can access and analyze accidents in real time, quickly locate the accident location, assess the severity, and formulate emergency guidance work orders in combination with public opinion factors to improve the accuracy and timeliness of accident handling, control the escalation of accidents, reduce the scope and duration of heating outages, and alleviate public opinion pressure. (4) The multi-source work order integration module of the present invention uses the work order identification big model to realize the parsing of multiple types of work orders, and combines the profile matching and dispatch optimization model to ensure accurate adaptation, improve the quality and efficiency of work order processing, and ensure the balance of personnel load. (5) The production management mobile device of the present invention supports the full-process digitalization of work order reception, execution trajectory monitoring and status feedback, realizes the traceability and controllability of work order processing, synchronizes work order progress in real time, and facilitates timely adjustments by management personnel; (6) The production evaluation and management module of this invention uses a multi-dimensional comprehensive evaluation index system to quantify the production management effect, locate the shortcomings in each link of work order processing, formulate targeted optimization strategies, and continuously improve the work order management mechanism through closed-loop iteration, thereby promoting the improvement of the production management level of the heating system.

[0017] Other features and advantages will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic block diagram of a digital production management platform for a heating system based on artificial intelligence, according to the present invention. Figure 2 A flowchart for the training work order method of this invention is provided; Figure 3 A flowchart for the inspection and maintenance work order method of this invention is provided; Figure 4 A flowchart for the emergency guidance work order method of this invention is provided; Figure 5 This is a flowchart of the work order dispatch optimization method of the present invention; Figure 6 This invention provides a flowchart for the comprehensive evaluation of production management and the formulation of work order processing optimization strategies. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions 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, 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.

[0022] like Figure 1 As shown, this embodiment provides a digital production management platform for a heating system based on artificial intelligence, which includes: The inspection management module is used to obtain relevant information about the heating system inspection, call the pre-set heating system anomaly prediction model, and formulate inspection work orders in combination with sudden emergency events in the production process. The inspection and maintenance management module is used to call the pre-set heating equipment defect detection model to determine whether the heating equipment has physical defects, and to analyze the severity of the physical defects and the scope of their impact on heating operation based on the detection results, and to formulate inspection and maintenance work orders. The emergency safety management module is used to acquire multimodal data from the accident site when an accident occurs in the heating system, analyze the severity of the accident, the location of the accident, the damage to the equipment, and formulate emergency guidance work orders in combination with social public opinion factors; The multi-source work order integration management module connects with the inspection management module, maintenance management module, and emergency safety management module. It is used to acquire multi-source work order integration information, including inspection work orders, maintenance work orders, and emergency guidance work orders. After that, it calls the pre-trained work order recognition model to obtain the work order content recognition result and sends it to the corresponding work order management personnel for review. Combining the pre-built work order profile, work order processing personnel profile, and current personnel status, it establishes a work order dispatch optimization model to obtain the optimal personnel information for executing each type of work order. The production management mobile device connects to the multi-source work order integration management module to receive and execute various types of work orders dispatched by personnel, monitor the trajectory information and work order processing status of the entire execution process of each type of work order, and provide feedback to the multi-source work order integration management module. The production evaluation management module connects with the multi-source work order integration management module to obtain the processing status of various types of work orders in the current heating season, set comprehensive evaluation indicators for multiple types, conduct a comprehensive evaluation of the production management situation in the heating season, analyze the mechanisms that can be optimized in the work order processing process, and formulate work order processing optimization strategies.

[0023] It should be noted that anomalies are defined as deviations in parameters, performance fluctuations, or signs of potential risks in heating equipment or systems during operation. These deviations have not yet resulted in substantial damage and can be addressed through planned interventions, such as adjusting operating strategies or increasing inspection frequency to identify the root cause and prevent defects or accidents. Defects are defined as physical damage, performance degradation, or functional abnormalities in heating equipment. These are potential faults that have not yet caused systemic damage and can be addressed through planned measures, such as inspecting corroded pipes and replacing aging equipment during the off-season, to eliminate potential defects, ensure normal equipment operation, and prevent accidents. Accidents are defined as sudden, destructive events caused by equipment failure, operational errors, or external factors, resulting in equipment damage, system paralysis, or large-scale heating outages. These accidents have a wide impact, potentially affecting multiple heat exchange stations or areas, and require immediate response to control their escalation, minimize losses, and manage public opinion. Anomalies are early signals of defects, and defects are potential causes of accidents. Timely intervention in anomalies can prevent defects from forming, and effective repair of defects can prevent accidents, forming a closed loop of risk control for anomalies, defects, and accidents.

[0024] like Figure 2 As shown, in this embodiment, the process of obtaining relevant information about the heating system inspection and calling a pre-set heating system anomaly prediction model, combined with sudden emergency events during the production process, to formulate an inspection work order includes: Obtain relevant information for heating system inspections, including physical information of heating equipment, operating data, and historical inspection records; Acquire information on sudden emergencies during the production process, including cold wave weather warnings and equipment malfunction alarms; Based on information related to heating system inspections and sudden emergencies in the production process, machine learning algorithms are used to extract abnormal features and train models to establish a heating system anomaly prediction model. A pre-set heating system anomaly prediction model is invoked to predict whether the heating system will experience potential anomalies. If no potential anomalies are found, a routine inspection work order is created. Otherwise, in addition to creating routine inspection work orders, a special inspection work order is created based on the type, location, and anomaly data of the potential anomalies. The inspection work order includes the inspection cycle, inspection route, and inspection tasks. The types of potential anomalies include abnormal equipment parameters and functional abnormalities during system operation, without causing substantial damage.

[0025] It should be noted that machine learning algorithms are used for anomaly feature extraction and model training to establish an anomaly prediction model for the heating system: the LSTM model is used to extract dynamic features of time series and cold wave scene features, and the random forest algorithm is used to calculate the importance of features and retain core features. Then, the XGBoost model is selected for model training to establish an anomaly prediction model for the heating system.

[0026] like Figure 3 As shown, in this embodiment, the step of calling a preset heating equipment defect detection model to determine whether the heating equipment has physical defects, and analyzing the severity of the physical defects and their impact on heating operation based on the detection results, and formulating a maintenance work order, includes: By calling the multi-source data acquisition submodule to collect basic information, operating data, historical defects and maintenance records of heating equipment, and by traversing and locating relevant technical information on heating maintenance and defect detection from network channels, a heating equipment defect database is constructed. Set up a prompt template for physical defect detection of heating equipment: clearly define the defect type, detection task, detection characteristics, and output defect detection result format; The database of heating equipment defects is input into a pre-trained large model. The model is then guided by prompt word templates to identify physical defects of heating equipment and extract defect feature classifications through semantic understanding, visual analysis and feature matching. The output of defect detection results includes whether physical defects have occurred, the location of the defect in the equipment, the description features of the defect type, the cause features of the defect and the range of the defect's impact. The large language model is used to perform semantic similarity analysis on the output defect features, and redundant features are clustered and merged. The defect detection results containing defect features are transformed into core input variables for defect risk modeling, and a defect risk model is established. Based on the defect risk model, risk values ​​are output, and risk levels are classified according to the risk values. When a defect is classified as high risk, its maintenance is defined as high priority and personnel are immediately assigned to handle it. When a defect is classified as medium or low risk, maintenance is scheduled during off-peak heating periods in conjunction with the heating production scheduling plan. Repair and maintenance work orders are formulated based on the equipment location of the defect, the characteristics of the defect type, the characteristics of the cause of the defect, the characteristics of the scope of the defect's impact, and the risk level.

[0027] The large-scale model used supports multimodal input of text and vision, such as the Tongyi Qianwen multimodal version, which has strong semantic understanding, visual analysis, and feature matching capabilities. Semantic understanding: Based on textual knowledge in the defect database, it parses device anomaly descriptions and historical records to match defect types; Visual analysis: It performs target detection and feature extraction on input images / videos to identify the shape, size, and color visual features of defect areas; Feature matching: It compares the parsed textual and visual features with standard defect features in the defect database to confirm the defect type and key attributes.

[0028] The semantic similarity calculation capability of a large language model (such as GPT-3.5 / 4) is selected as the input, along with the text containing all defect features to be clustered. Semantic similarity scores between features are calculated (using the cosine similarity algorithm). A similarity threshold is set to determine feature redundancy. A hierarchical clustering algorithm is used to group defect features with semantic similarity higher than the threshold into one category. For each category of redundant features, core common information is extracted to generate unified standardized features. A random forest model is used for training and learning the defect risk model.

[0029] In this embodiment, the basic information of the heating equipment includes at least the model, installation time, relevant parameters, and maintenance records of the heat source unit, heat exchanger, heating network pipeline, pumps, and valves; the operating data of the heating equipment includes at least the equipment's start-up and shutdown status, load changes, heat metering parameters, and vibration parameters; and the historical defects and maintenance records of the heating equipment include at least the type of physical defect, the cause of the defect, the influencing factors of the defect, the handling method, and the repair effect. The types of physical defects in the equipment include at least the following: burner nozzle blockage in the heat source unit; corrosion or leakage in the fuel supply pipeline; wear and abnormal noise in auxiliary equipment; perforation of the heat exchanger tube wall; loose parts; scaling of the heat exchanger tube; blockage of the heat exchange medium channel; seal failure; corrosion of the inner and outer walls of the heating network pipeline; mechanical damage to the pipeline; detachment or damage of the insulation layer; pump body wear; bearing damage; internal leakage of valves; valve jamming; valve body cracks or weld leakage.

[0030] In this embodiment, the defect risk model is represented as: ; This represents the risk value for the i-th type of heating defect. This represents the probability of a defect occurring. The severity of the loss caused by the defect; Let be the probability density function of the probability of defect occurrence; Let be the probability density function of the severity of the loss.

[0031] In this embodiment, the maintenance management module is also used to predict the demand for maintenance materials based on historical maintenance data, equipment operating status, and maintenance work orders, using a combination of time series analysis algorithm and association rule mining algorithm, to optimize material inventory management and procure and reserve spare parts resources.

[0032] It should be noted that historical maintenance data includes: maintenance work orders from recent years (including spare parts models, consumption quantities, and replacement frequency), defect types, and corresponding material consumption records; equipment operation data includes: equipment operating years, cumulative operating time, load rate, key parameters, and anomaly frequency; and maintenance work orders include: defect types, number of equipment involved, and estimated material requirements in pending maintenance work orders. The system uses ARIMA time series analysis to capture the time trend and seasonal fluctuations in spare parts consumption, and employs association rule mining algorithms to identify the correlation between defect types and spare parts combinations, and between equipment status and spare parts consumption. This comprehensive prediction of maintenance material requirements is then integrated with the inventory management system to synchronize spare parts inventory quantities and in-transit procurement quantities in real time. When spare parts inventory falls below the safety stock level, an early warning is automatically triggered, and procurement timing and quantity are planned.

[0033] like Figure 4 As shown, in this embodiment, when an accident occurs, the heating system acquires multimodal data from the accident site, analyzes the severity of the accident, its location, and the extent of equipment damage, and, in conjunction with public opinion factors, formulates an emergency guidance work order, including: When a sudden, destructive accident occurs in the heating system due to equipment failure, operational error, or external factors, relevant data caused by the accident will be obtained, including historical accident data, real-time abnormal system data, images of equipment damage caused by the accident, geographic information, and heating accident reports from social media and news platforms. After retrieving relevant accident entities from the acquired data and standards, emergency plans, and risk prevention manuals based on the heating system using a multimodal large model, the causal relationship between the accident entities, the temporal relationship between the accident and time, the locational relationship between the accident and geography, and the relationship between the accident and emergency repair measures are determined. Then, a knowledge graph of heating accidents is constructed with accident entities as nodes and relationships as edges. Based on the knowledge graph of heating accidents, accident chain nodes are identified, including the cause node that triggers the accident, the accident occurrence node, the diffusion node that triggers secondary accidents, and the accident response node. At the same time, the impact of the accident on the heating area, the duration of the heating outage, and the public opinion intensity are analyzed, and emergency guidance work orders are formulated, including emergency repair equipment, emergency repair location, personnel allocation, and public opinion response.

[0034] It should be noted that a large model supporting multimodal input of text, image, and voice is used, combined with retrieval-enhanced generation technology, to extract core accident entities through semantic understanding and visual analysis. These entities include: accident-related entities, cause-related entities, accident-causing-impact-related entities, location-related entities, measure-related entities, and time-related entities. Semantic analysis of the large model identifies entity relationships: causal relationships (e.g., cause-accident, accident-secondary accident), temporal relationships (e.g., accident occurrence time, duration correlation), locational relationships (e.g., location of the heat exchange station or community where the accident occurred), and response relationship (e.g., the correspondence between accident and response measures, and between measures and tools).

[0035] Using accident entities as nodes and four types of relationships as directed edges, a knowledge graph is constructed using the Neo4j graph database. The knowledge graph is traversed using graph algorithms (such as shortest path and centrality analysis) to identify key accident chain nodes: cause nodes: initial factors that trigger accidents; occurrence nodes: core accident events; diffusion nodes: key nodes that trigger secondary accidents; and response nodes: matching emergency response measures.

[0036] By combining knowledge graph-related data with real-time data, the impact of accidents is quantified: impact on heating area: calculated by the pipeline coverage area associated with location nodes; impact on heating outage duration: predicted based on historical data on similar accidents and the matching degree of current measures; impact on public opinion intensity: predicted intensity level by extracting keywords from public opinion data and combining them with historical public opinion dissemination models.

[0037] like Figure 5 As shown, in this embodiment, after acquiring multi-source work order integrated information including inspection work orders, maintenance work orders, and emergency guidance work orders, a pre-trained work order recognition model is invoked to obtain the work order content recognition result, which is then sent to the corresponding work order management personnel for review. Combining the pre-built work order profile, work order processing personnel profile, and current personnel status, a work order dispatch optimization model is established to obtain the optimal personnel information for executing each type of work order, including: Acquire integrated work order information from multiple sources, including inspection work orders, maintenance work orders, and emergency guidance work orders, with at least one work order of each type; After pre-training and fine-tuning the work order recognition model using historical multi-source work order integrated information, the acquired multi-source work order integrated information is input into the work order recognition model. The work order content recognition results, including work order type tags, heating area tags involved in the work order, core business content tags of the work order, and work order priority tags, are identified through semantic understanding and then sent to the corresponding work order management personnel for manual review. After manual review and approval, the work order dispatch submodule is entered to obtain the responsiveness requirements, personnel skill and qualification requirements, and special scenario requirements of each work order as work order service preference features. The content recognition results of each work order are used as the basic features of the work order. Based on the basic features and service preference features of the work order, a profile of each work order is constructed. The basic characteristics and service capability characteristics of each work order processing personnel are obtained to construct a profile of each work order processing personnel. The basic characteristics of the work order processing personnel include skill tags, qualification level tags, current work status, and historical work order processing records. The service capability characteristics of the work order processing personnel include the completion rate of the personnel's work orders, the response speed of the personnel's work orders, the matching degree of the personnel's skill certification, and the personnel's experience in handling special scenarios. Set up feature mapping matching rules between work order profiles and work order processing personnel profiles, perform preliminary quantitative matching between work orders and personnel to form a candidate pool of personnel for each work order processing, and analyze the personnel matching degree by combining the calculated feature weights and feature matching scores of each dimension. Through optimization algorithm, select the personnel with the highest matching degree and balanced work order load rate from the candidate pool as the optimal personnel for work order processing.

[0038] It should be noted that the feature mapping matching rules between the work order profile and the work order processing personnel profile are set as follows: Based on the core feature dimensions of the work order profile and the personnel profile, corresponding mapping rules are established, including: There are mapping rules between the work order type and skill qualification requirements in the work order profile feature dimension and the skill tags, qualification levels and skill certification matching degree in the personnel profile feature dimension; There is a mapping rule between the responsiveness requirements in the work order profile feature dimension and the response speed in the personnel profile feature dimension; There is a mapping rule between the special scenario requirements in the work order profile feature dimension and the special scenario handling experience in the personnel profile feature dimension; There are mapping rules between the work order profile feature dimension, which includes the equipment and task complexity of the work order, and the personnel profile feature dimension, which includes the historical work order processing records, the completion rate of personnel's work orders, and the current work status.

[0039] Preliminary quantitative matching forms a candidate pool of personnel to handle each work order: For each mapping rule dimension, a single-dimensional score is given according to the preset criteria of fully compliant matching, basically compliant matching, partially compliant matching, and non-compliant matching. The score is set to be greater than the total matching score threshold as the admission condition. After traversing all personnel, the personnel who meet the condition are selected as the candidate pool of personnel to handle the work order, and they are arranged in descending order of score to form a priority list of candidate pools.

[0040] The matching degree of personnel is analyzed by combining the calculated feature weights and matching scores of each dimension. The optimal personnel for work order processing is selected from the candidate pool by an optimization algorithm, based on the highest matching degree and the balanced work order load rate. The AHP method is used to compare the importance of each mapping rule matching dimension pairwise to construct a judgment matrix. The feature vector is calculated by matrix operation, and the feature weights of each dimension are determined after consistency verification. Based on the single-dimensional matching scores and weights of the personnel in the candidate pool, the comprehensive matching degree of each personnel is calculated. The primary objective function is to achieve the highest comprehensive matching degree score, and the secondary objective function is to balance the work order load rate of personnel. At the same time, personnel load rate constraints and time limit constraints are set. The optimal personnel that meet the objective functions are selected from the candidate pool by a weighted objective programming algorithm.

[0041] like Figure 6 As shown, in this embodiment, the process of obtaining the processing status of various types of work orders during the current heating season, setting comprehensive evaluation indicators for multiple types, comprehensively evaluating the production management situation during the heating season, analyzing the mechanisms that can be optimized in the work order processing process, and formulating work order processing optimization strategies includes: Acquire full-process data on the creation, response, processing, and feedback of various types of work orders during the current heating season, and set up a comprehensive evaluation of multiple indicators, including work order creation compliance, work order processing efficiency, processing quality, resource utilization efficiency, and risk prevention and control capabilities. The weights of each type of indicator are determined by the analytic hierarchy process (AHP). The scores of each type of indicator are combined and weighted to obtain the total comprehensive evaluation score. Based on the total comprehensive evaluation score, the production management situation during the heating season is classified into different levels. At the same time, a production management evaluation report for the heating season is generated, including the comprehensive evaluation score, the scores of each indicator, and the labeling of the weakest indicators. Based on the identified shortcomings, a root cause analysis was conducted to identify the key factors affecting the indicator scores, including non-compliant work order content and unreasonable personnel assignment, and corresponding work order processing optimization strategies were developed.

[0042] In practical applications, analyzing work order processing data helps identify bottlenecks and inefficient areas in the process. If a region has a low work order response rate, further analysis can be conducted to determine if it's due to insufficient staffing, inefficient scheduling, or equipment malfunctions. Through this in-depth analysis, companies can take targeted measures to optimize staffing, improve scheduling strategies, or strengthen equipment maintenance, thereby enhancing overall heating production management.

[0043] In this embodiment, the compliance indicators for work order formulation include work order element completeness rate, work order classification accuracy rate, and work order priority matching rate; the work order processing efficiency indicators include average response time, response timeliness rate, average processing time, and overdue completion rate; the processing quality indicators include acceptance pass rate, defect / accident recurrence rate, and work order acceptance score; the resource utilization efficiency indicators include personnel load balance, average work order processing volume per person, and material consumption deviation rate; and the risk prevention and control capability indicators include defect identification accuracy rate, defect-accident conversion rate, emergency response success rate, and major accident occurrence rate.

[0044] In practical applications, the comprehensive evaluation score calculation and work order processing optimization process includes: 1) The weights of each secondary indicator are calculated using the Analytic Hierarchy Process (AHP); (the primary indicators are work order formulation compliance indicators, work order processing efficiency indicators, processing quality indicators, resource utilization efficiency indicators, and risk control capability indicators; the secondary indicators are the specific content of each primary indicator). 2) Calculate the raw scores for each secondary indicator, and convert the raw scores into standard scores from 0 to 100. For example: The work order element completeness rate is equal to the number of work orders with complete elements divided by the total number of work orders. For example, if there are 100 work orders and 85 have complete elements, the score is 85 points. The average response time is equal to the sum of the response times of each individual work order divided by the number of work orders. For example, if there are 3 work orders with response times of 20, 30, and 25 minutes respectively, the average response time is 25 minutes. This is then converted into a standard score based on the regulations. For example, if the regulation stipulates a response time within 30 minutes, a score of 25 minutes can be counted as 90 points. 3) Calculate the scores of the primary indicators and the total comprehensive evaluation score by weighted summation; 4) Evaluation result classification: Excellent: A score greater than 90 indicates that the heating system's production management process is smooth, and the work order creation, response, and processing are efficient and of high quality. Good: 80-89 points, indicating that the overall performance meets the standard, but some aspects can be improved; Pass: 60-79 points, indicating that the core indicators meet the standards, but there are still some shortcomings that can be optimized; Unqualified: A score of less than 60 indicates serious problems in the production management of the heating system, requiring urgent optimization. 5) Tracing the root causes of shortcomings: For example, low response timeliness can be analyzed to determine whether it is due to unreasonable personnel distribution, skill mismatch, or a lagging scheduling mechanism. Furthermore, by using correlation algorithms, key factors affecting the overall score can be identified. 6) Typical optimization strategies: Efficiency optimization: To address the low response time rate, optimize the allocation of personnel to different areas; to address the excessively long average processing time, conduct specialized skills training. Quality optimization: To address the high defect recurrence rate, we updated the standard maintenance process and added a post-repair re-inspection step; to address the low acceptance rate, we optimized the work order-personnel matching model. Resource optimization: To address uneven staff workload, adjust work order dispatch rules to avoid oversaturation of core personnel; to address unreasonable material consumption, optimize material demand forecasting models and establish dynamic inventory early warning systems. Risk optimization: To address the low accuracy of defect identification, upgrade the inspection model and increase the inspection frequency of key equipment; to address the high accident conversion rate, establish a defect classification and handling mechanism, prioritizing the dispatch of high-risk defects.

[0045] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0046] Furthermore, the functional modules in the various embodiments of this invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part. If the function is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0047] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A digital production management platform for heating systems based on artificial intelligence, characterized in that, It includes: The inspection management module is used to obtain relevant information about the heating system inspection, call the pre-set heating system anomaly prediction model, and formulate inspection work orders in combination with sudden emergency events in the production process. The inspection and maintenance management module is used to call the pre-set heating equipment defect detection model to determine whether the heating equipment has physical defects, and to analyze the severity of the physical defects and the scope of their impact on heating operation based on the detection results, and to formulate inspection and maintenance work orders. The emergency safety management module is used to acquire multimodal data from the accident site when an accident occurs in the heating system, analyze the severity of the accident, the location of the accident, the damage to the equipment, and formulate emergency guidance work orders in combination with social public opinion factors; The multi-source work order integration management module connects with the inspection management module, maintenance management module, and emergency safety management module. It is used to acquire multi-source work order integration information, including inspection work orders, maintenance work orders, and emergency guidance work orders. After that, it calls the pre-trained work order recognition model to obtain the work order content recognition result and sends it to the corresponding work order management personnel for review. Combining the pre-built work order profile, work order processing personnel profile, and current personnel status, it establishes a work order dispatch optimization model to obtain the optimal personnel information for executing each type of work order. The production management mobile device connects to the multi-source work order integration management module to receive and execute various types of work orders dispatched by personnel, monitor the trajectory information and work order processing status of the entire execution process of each type of work order, and provide feedback to the multi-source work order integration management module. The production evaluation management module connects with the multi-source work order integration management module to obtain the processing status of various types of work orders in the current heating season, set comprehensive evaluation indicators for multiple types, conduct a comprehensive evaluation of the production management situation in the heating season, analyze the mechanisms that can be optimized in the work order processing process, and formulate work order processing optimization strategies.

2. The digital production management platform for heating systems according to claim 1, characterized in that, The process involves acquiring relevant information about the heating system inspection, calling a pre-set heating system anomaly prediction model, and combining this with unexpected emergencies during the production process to create inspection work orders, including: Obtain relevant information for heating system inspections, including physical information of heating equipment, operating data, and historical inspection records; Acquire information on sudden emergencies during the production process, including cold wave weather warnings and equipment malfunction alarms; Based on information related to heating system inspections and sudden emergencies in the production process, machine learning algorithms are used to extract abnormal features and train models to establish a heating system anomaly prediction model. A pre-set heating system anomaly prediction model is invoked to predict whether the heating system will experience potential anomalies. If no potential anomalies are found, a routine inspection work order is created. Otherwise, in addition to creating routine inspection work orders, a special inspection work order is created based on the type, location, and anomaly data of the potential anomalies. The inspection work order includes the inspection cycle, inspection route, and inspection tasks. The types of potential anomalies include abnormal equipment parameters and functional abnormalities during system operation, without causing substantial damage.

3. The digital production management platform for heating systems according to claim 1, characterized in that, The process involves calling a pre-set heating equipment defect detection model to determine whether the heating equipment has physical defects, analyzing the severity of the physical defects and their impact on heating operation based on the detection results, and then creating a maintenance work order, including: By calling the multi-source data acquisition submodule to collect basic information, operating data, historical defects and maintenance records of heating equipment, and by traversing and locating relevant technical information on heating maintenance and defect detection from network channels, a heating equipment defect database is constructed. Set up a prompt template for physical defect detection of heating equipment: clearly define the defect type, detection task, detection characteristics, and output defect detection result format; The database of heating equipment defects is input into a pre-trained large model. The model is then guided by prompt word templates to identify physical defects of heating equipment and extract defect feature classifications through semantic understanding, visual analysis and feature matching. The output of defect detection results includes whether physical defects have occurred, the location of the defects in the equipment, the description features of the defect type, the cause features of the defects, and the range of the defects. The large language model is used to perform semantic similarity analysis on the output defect features, and redundant features are clustered and merged. The defect detection results containing defect features are transformed into core input variables for defect risk modeling, and a defect risk model is established. Based on the defect risk model, risk values ​​are output, and risk levels are classified according to the risk values. When a defect is classified as high risk, its maintenance is defined as high priority and personnel are immediately assigned to handle it. When a defect is classified as medium or low risk, maintenance is arranged during off-peak heating periods in conjunction with the heating production scheduling plan. Repair and maintenance work orders are formulated based on the equipment location of the defect, the characteristics of the defect type, the characteristics of the cause of the defect, the characteristics of the scope of the defect's impact, and the risk level.

4. The digital production management platform for heating systems according to claim 3, characterized in that, The basic information of the heating equipment includes at least the model, installation time, relevant parameters, and maintenance records of the heat source unit, heat exchanger, heating network pipeline, pumps and valves; the operating data of the heating equipment includes at least the equipment start-up and shutdown status, load changes, heat metering parameters, and vibration parameters; the historical defects and maintenance records of the heating equipment include at least the type of physical defect, the cause of the defect, the influencing factors of the defect, the handling method, and the repair effect. The types of physical defects in the equipment include at least the following: burner nozzle blockage in the heat source unit; corrosion or leakage in the fuel supply pipeline; wear and abnormal noise in auxiliary equipment; perforation of the heat exchanger tube wall; loose parts; scaling of the heat exchanger tube; blockage of the heat exchange medium channel; seal failure; corrosion of the inner and outer walls of the heating network pipeline; mechanical damage to the pipeline; detachment or damage of the insulation layer; pump body wear; bearing damage; internal leakage of valves; valve jamming; valve body cracks or weld leakage.

5. The digital production management platform for heating systems according to claim 3, characterized in that, The defect risk model is expressed as follows: ； This represents the risk value for the i-th type of heating defect. This represents the probability of a defect occurring. The severity of the loss caused by the defect; Let be the probability density function of the probability of defect occurrence; Let be the probability density function of the severity of the loss.

6. The digital production management platform for heating systems according to claim 1, characterized in that, The maintenance management module is also used to predict maintenance material demand based on historical maintenance data, equipment operating status, and maintenance work orders, using a combination of time series analysis and association rule mining algorithms, to optimize material inventory management and procure and reserve spare parts resources.

7. The digital production management platform for heating systems according to claim 1, characterized in that, When a heating system accident occurs, it acquires multimodal data from the accident site, analyzes the severity of the accident, its location, and the extent of equipment damage, and, in conjunction with public opinion factors, formulates emergency guidance work orders, including: When a sudden, destructive accident occurs in the heating system due to equipment failure, operational error, or external factors, relevant data caused by the accident will be obtained, including historical accident data, real-time abnormal system data, images of equipment damage caused by the accident, geographic information, and heating accident reports from social media and news platforms. After retrieving relevant accident entities from the acquired data and standards, emergency plans, and risk prevention manuals based on the heating system using a multimodal large model, the causal relationship between the accident entities, the temporal relationship between the accident and time, the locational relationship between the accident and geography, and the relationship between the accident and emergency repair measures are determined. Then, a knowledge graph of heating accidents is constructed with accident entities as nodes and relationships as edges. Based on the knowledge graph of heating accidents, accident chain nodes are identified, including the cause node that triggers the accident, the accident occurrence node, the diffusion node that triggers secondary accidents, and the accident response node. At the same time, the impact of the accident on the heating area, the duration of the heating outage, and the public opinion intensity are analyzed, and emergency guidance work orders are formulated, including emergency repair equipment, emergency repair location, personnel allocation, and public opinion response.

8. The digital production management platform for heating systems according to claim 1, characterized in that, After acquiring multi-source work order integrated information, including inspection work orders, maintenance work orders, and emergency guidance work orders, a pre-trained work order recognition model is invoked to obtain work order content recognition results, which are then sent to the corresponding work order management personnel for review. Combining pre-built work order profiles, work order processing personnel profiles, and the current personnel status, a work order dispatch optimization model is established to obtain the optimal personnel information for executing each type of work order, including: Acquire integrated work order information from multiple sources, including inspection work orders, maintenance work orders, and emergency guidance work orders, with at least one work order of each type; After pre-training and fine-tuning the work order recognition model using historical multi-source work order integrated information, the acquired multi-source work order integrated information is input into the work order recognition model. The work order content recognition results, including work order type tags, heating area tags involved in the work order, core business content tags of the work order, and work order priority tags, are identified through semantic understanding and then sent to the corresponding work order management personnel for manual review. After manual review and approval, the work order dispatch submodule is entered to obtain the responsiveness requirements, personnel skill and qualification requirements, and special scenario requirements of each work order as work order service preference features. The content recognition results of each work order are used as the basic features of the work order. Based on the basic features and service preference features of the work order, a profile of each work order is constructed. The basic characteristics and service capability characteristics of each work order processing personnel are obtained to construct a profile of each work order processing personnel. The basic characteristics of the work order processing personnel include skill tags, qualification level tags, current work status, and historical work order processing records. The service capability characteristics of the work order processing personnel include the completion rate of the personnel's work orders, the response speed of the personnel's work orders, the matching degree of the personnel's skill certification, and the personnel's experience in handling special scenarios. Set up feature mapping matching rules between work order profiles and work order processing personnel profiles, perform preliminary quantitative matching between work orders and personnel to form a candidate pool of personnel for each work order processing, and analyze the personnel matching degree by combining the calculated feature weights and feature matching scores of each dimension. Through optimization algorithm, select the personnel with the highest matching degree and balanced work order load rate from the candidate pool as the optimal personnel for work order processing.

9. The digital production management platform for heating systems according to claim 1, characterized in that, The process involves acquiring the processing status of various types of work orders during the current heating season, setting comprehensive evaluation indicators for multiple types, conducting a comprehensive evaluation of the production management situation during the heating season, analyzing the mechanisms that can be optimized in the work order processing process, and formulating work order processing optimization strategies, including: Acquire full-process data on the creation, response, processing, and feedback of various types of work orders during the current heating season, and set up a comprehensive evaluation of multiple indicators, including work order creation compliance, work order processing efficiency, processing quality, resource utilization efficiency, and risk prevention and control capabilities. The weights of each type of indicator are determined by the analytic hierarchy process (AHP). The scores of each type of indicator are combined and weighted to obtain the total comprehensive evaluation score. Based on the total comprehensive evaluation score, the production management situation during the heating season is classified into different levels. At the same time, a production management evaluation report for the heating season is generated, including the comprehensive evaluation score, the scores of each indicator, and the labeling of the weakest indicators. Based on the identified shortcomings, a root cause analysis was conducted to identify the key factors affecting the indicator scores, including non-compliant work order content and unreasonable personnel assignment, and corresponding work order processing optimization strategies were developed.

10. The digital production management platform for heating systems according to claim 1, characterized in that, The compliance indicators for work order formulation include work order element completeness rate, work order classification accuracy rate, and work order priority matching rate; the work order processing efficiency indicators include average response time, response timeliness rate, average processing time, and overdue completion rate; the processing quality indicators include acceptance pass rate, defect / accident recurrence rate, and work order acceptance score; the resource utilization efficiency indicators include personnel load balance, average work order processing volume per person, and material consumption deviation rate; and the risk prevention and control capability indicators include defect identification accuracy rate, defect-accident conversion rate, emergency response success rate, and major accident occurrence rate.

Images