Environmental protection design scheme generation method and system

Abstract

The invention discloses an environmental protection design scheme generation method and system, particularly relates to the technical field of environmental protection design optimization, and is used for solving the technical problem that an existing environmental protection design scheme is disjointed with real-time production requirements due to data updating lag. The method comprises the following steps: screening incremental data based on a timestamp sequence and predicting a dynamic environmental protection constraint threshold range, identifying a supply chain mutation node through a carbon emission trend vector and marking an optimization region, generating a candidate scheme set in combination with a business logic association rule, and eliminating an invalid scheme through dual verification of production feasibility and historical version compatibility. And finally, mapping the remaining schemes to a dynamic threshold range and outputting an optimization result. The data dynamic change can be actively responded, the scheme landing efficiency is improved on the premise of guaranteeing environmental protection indexes, meanwhile, the stability of the scheme in long-term data iteration is ensured through cross-version compatibility verification, and resource waste caused by design reworking is effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental protection design optimization. More specifically, the present invention relates to a method and system for generating an environmental protection design solution. Background Art

[0002] In the prior art, the environmental protection design method based on life cycle assessment (LCA) usually relies on a static or periodically updated database, such as material carbon emission data, energy consumption indicators, etc. These data are collected through a standardized process and integrated into the design system to optimize the environmental performance of the entire life cycle of the product. However, the update cycle of LCA data often has a significant difference from the real-time requirements of design optimization. For example, the design system needs to generate a solution based on the latest data, but in actual engineering, the update frequency of the LCA database is restricted by the complexity of supply chain data collection, industry collaboration efficiency, and local data model construction, resulting in the difficulty of the design optimization process to dynamically respond to data changes.

[0003] In the prior art, there is a contradiction between the dynamic lag of LCA data and the real-time nature of design optimization. That is, since environmental protection design needs to rely on multi-source heterogeneous data (such as material production, supply chain carbon footprint, etc.), and the data update cycle is long and the collaboration mechanism is insufficient, the data used by the design system in the optimization process is disconnected from the environmental protection indicators at the actual production stage. For example, the design solution is generated based on historical data, but in the actual production stage, due to data update, the carbon emission exceeds the standard, and the process or material selection needs to be readjusted, resulting in waste of resources and loss of efficiency. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a method and system for generating an environmental protection design solution to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for generating an environmental protection design solution, comprising the following steps: S1. Obtain the life cycle assessment data of the target product and its time stamp sequence, and the life cycle assessment data includes the carbon emission intensity of the supply chain node; S2. Screen the incremental data updated in the last preset number of cycles according to the time stamp sequence, and predict the range of the next cycle's dynamic environmental protection constraint threshold based on the historical constraint change trajectory to determine whether the incremental data deviates; S3. If it deviates, construct a time series carbon emission trend vector according to the historical data of the supply chain node, and mark the mutation node exceeding the preset threshold in the carbon emission trend vector as the optimization area; S4. Generate a candidate solution set based on the carbon emission intensity of the optimization area and the business logic association rule of the mutation node; S5. Conduct production feasibility verification on the candidate solution set, and perform reverse compatibility verification between the candidate solutions and the historical version life cycle assessment data to eliminate invalid solutions; S6. Map the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range, and output the environmental protection design solution set.

[0006] In a preferred implementation, obtain the life cycle assessment data of the target product and its time stamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain nodes, including: Obtain the carbon emissions corresponding to the raw material production stage of the supply chain node and the energy consumption types corresponding to the transportation stage. Match the energy consumption types with the preset carbon emission coefficient table to obtain the carbon emission intensity of the transportation stage; Classify the data of the supply chain nodes according to the preset industry classification standard, and bind the classified data to the corresponding storage path; Generate a time stamp sequence according to the regional location of the data acquisition terminal, where the accuracy of the time stamp sequence is associated with the data update frequency.

[0007] In a preferred implementation, screen the incremental data updated in the last preset number of cycles according to the time stamp sequence, and predict the dynamic environmental protection constraint threshold range in the next cycle based on the historical constraint change trajectory to determine whether the incremental data deviates, including: Screen out the incremental data updated in the last preset number of cycles according to the time order of the time stamp sequence. The incremental data updated in the last preset number of cycles is the first preset items of data intercepted after sorting from new to old according to the time stamp; Extract the change data of the dynamic environmental protection constraint threshold in the historical cycle, construct the threshold change trend within the time window according to the change data, and calculate the dynamic environmental protection constraint threshold range in the next cycle through the threshold change trend; Compare the carbon emission intensity in the incremental data with the upper and lower limit values of the dynamic environmental protection constraint threshold range. If it exceeds the upper limit value or is lower than the lower limit value, it is determined to deviate.

[0008] In a preferred implementation, if there is a deviation, construct a carbon emission trend vector of the time series according to the historical data of the supply chain node, and mark the mutation nodes exceeding the preset threshold in the carbon emission trend vector as the optimization area, including: Extract the carbon emission intensity data of the supply chain node in the historical cycle, arrange it in time order to generate a carbon emission intensity sequence, perform sliding window processing on the carbon emission intensity sequence, calculate the mean value and volatility of the carbon emission intensity in each window, and combine the mean value and volatility into a multi-dimensional trend vector; Calculate the deviation value between the average carbon emission intensity of each node in the multi-dimensional trend vector and the current dynamic environmental protection constraint threshold range. If the deviation value exceeds the preset fluctuation threshold, it is determined as a mutation node, and the mutation node and its associated supply chain nodes are marked as the optimization area; The marking rule for the optimization area is to expand the marking range according to the business relevance of supply chain nodes. If there is a contract binding or geographical proximity relationship between the mutation node and the associated node, the associated node is synchronously added to the optimization area.

[0009] In a preferred embodiment, based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation node, a set of candidate solutions is generated, including: Obtain the carbon emission intensity data of the mutation node in the optimization area and the business logic rules of the associated supply chain nodes. Extract the upstream and downstream nodes with direct supply and demand association with the mutation node through the contract binding relationship to form a supply chain carbon flow impact chain; Based on the supply chain carbon flow impact chain, analyze the superposition effect of alternative nodes on the overall supply chain carbon emissions, and generate a set of alternative solutions. Each alternative solution includes the predicted values of the carbon emission intensity changes of the alternative node and the corresponding upstream and downstream nodes; According to the geographical proximity relationship, screen the node combinations that can merge transportation routes, and generate a set of collaborative solutions in combination with the path priority stratification mechanism. The priority stratification mechanism is to optimize the transportation paths with high delay sensitivity first; Multidimensionally weighted and fuse the set of alternative solutions and the set of collaborative solutions based on the carbon emission intensity threshold and path priority, and eliminate the solutions whose superposition effect exceeds the preset risk threshold to generate a set of candidate solutions; Perform a pre-check on the process compatibility conflicts for each solution in the set of candidate solutions.

[0010] In a preferred embodiment, the pre-check rule is to match the production parameter range of the alternative node with the limit value of the current process equipment, and the pre-check result of the conflict triggers the solution dynamic adjustment mechanism.

[0011] In a preferred embodiment, perform production feasibility verification on the set of candidate solutions, and perform reverse compatibility verification on the candidate solutions and the historical version life cycle assessment data, and eliminate the invalid solutions, including: Extract the process parameters and material property data in the set of candidate solutions, compare them with the compatibility database of the current production equipment, and screen the solutions with process parameter deviation values lower than the preset threshold; Input the screened candidate solutions into the historical version life cycle assessment model, and verify one by one whether their environmental protection scores in the historical version data continue to meet the standards; If the candidate solution passes the current production equipment compatibility verification but fails the historical version environmental protection score verification, it is determined as an invalid solution and eliminated; Mark the failure causes of the failure solutions. The failure causes include process parameter conflicts and excessive fluctuations in historical version scores. Store the marking results in the failure solution database for subsequent optimization reference.

[0012] In a preferred embodiment, the verification method for verifying whether the environmental protection score in the historical version data is continuously qualified is to substitute the scheme parameters into the calculation engine of the corresponding historical version for re-scoring.

[0013] In a preferred embodiment, map the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range, and output a set of environmental protection design solutions, including: Obtain the upper and lower limit values of the predicted dynamic environmental protection constraint threshold range for the next period; Calculate the normalized percentage of the comprehensive environmental protection scores of the remaining candidate solutions within the dynamic environmental protection constraint threshold range; Screen the candidate solutions with the normalized percentage within the preset acceptable interval; Sort the screened candidate solutions from high to low according to the normalized percentage to generate a preliminary set of environmental protection design solutions; Perform weighted correction on the preliminary set of environmental protection design solutions according to the stability score of the historical version compatibility verification. The higher the stability score, the greater the sorting weight, and generate the final set of environmental protection design solutions.

[0014] On the other hand, the present invention provides an environmental protection design solution generation system, including the following modules: A data acquisition module for acquiring the life cycle assessment data of the target product and its time stamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain nodes; A dynamic prediction module for screening the incremental data updated in the last preset number of cycles according to the time stamp sequence, and predicting the dynamic environmental protection constraint threshold range for the next cycle based on the historical constraint change trajectory to determine whether the incremental data deviates; If it deviates, the optimization marking module constructs a time series carbon emission trend vector based on the historical data of the supply chain nodes, and marks the mutation nodes exceeding the preset threshold in the carbon emission trend vector as the optimization area; A solution generation module for generating a set of candidate solutions based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes; A verification and elimination module for verifying the production feasibility of the set of candidate solutions, and performing reverse compatibility verification on the candidate solutions and the historical version life cycle assessment data, and eliminating the failure solutions; A result output module for mapping the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range and outputting a set of environmental protection design solutions.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. By constructing a dynamic data-driven mechanism and a multi-dimensional verification system, the problem of disconnection of solutions caused by data lag in traditional environmental protection designs is effectively solved, significantly improving the real-time performance of design optimization and the implementation reliability of solutions. By integrating the dynamic threshold prediction of the timestamp sequence and the intelligent marking of the optimization area, mutation nodes in the supply chain can be quickly identified within the data update cycle, and candidate solutions covering upstream and downstream linkages can be generated in combination with business rules, thereby minimizing the risk of out-of-control carbon emissions across the entire chain caused by local data fluctuations while ensuring environmental protection indicators. At the same time, based on the dual screening mechanism of historical version compatibility verification and production feasibility, it is ensured that the final solution not only adapts to current environmental protection constraints but also maintains stability during long-term data iteration, avoiding waste of resources caused by frequent rework.

[0016] 2. By deeply coupling dynamic prediction, intelligent marking, and multi-dimensional verification, a closed-loop feedback environmental protection design system is formed. By real-time tracking the changes in incremental data and predicting the future threshold range, the design solution always conforms to industry standards. On the other hand, the carbon emission trend vector is used to accurately locate the optimization area and associate business rules to ensure that optimization measures can penetrate all links of the supply chain and achieve global carbon emission control in scenarios such as material replacement and process adjustment. In addition, the reverse compatibility verification mechanism further strengthens the robustness of the solution through cross-version data backtracking, enabling it to maintain high availability and low maintenance costs in complex engineering environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flowchart of a method for generating an environmental protection design solution of the present invention; Figure 2 is a schematic structural diagram of a system for generating an environmental protection design solution of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] Embodiment 1: Figure 1 A method for generating an environmental protection design solution of the present invention is given, which includes the following steps: S1. Obtain the life cycle assessment data of the target product and its timestamp sequence, where the life cycle assessment data includes the carbon emission intensity of supply chain nodes; S2. Screen the incremental data updated in the last preset number of cycles according to the timestamp sequence, predict the range of the dynamic environmental protection constraint threshold for the next cycle based on the historical constraint change trajectory, and determine whether the incremental data deviates; S3. If it deviates, construct a carbon emission trend vector of the time series based on the historical data of the supply chain node, and mark the mutation nodes exceeding the preset threshold in the carbon emission trend vector as the optimization area; S4. Generate a set of candidate solutions based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes; S5. Conduct production feasibility verification on the set of candidate solutions, and conduct reverse compatibility verification between the candidate solutions and the life cycle assessment data of the historical version, and eliminate the invalid solutions; S6. Map the comprehensive environmental protection scores of the remaining candidate solutions to the range of the dynamic environmental protection constraint threshold, and output the set of environmental protection design solutions.

[0020] S1. Obtain the life cycle assessment data of the target product and its timestamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain node. The specific implementation is as follows: Obtain the carbon emissions corresponding to the raw material production stage of the supply chain node and the types of energy consumption corresponding to the transportation stage. The types of energy consumption include diesel, electricity or natural gas consumed by transportation tools. The carbon emissions in the raw material production stage are collected through the production equipment monitoring system connected to the supply chain node. The production equipment monitoring system records the power consumption value, fuel consumption value and the corresponding production batch time range during the production process. The types of energy consumption in the transportation stage are obtained through the transportation records of the logistics management system. The transportation records include the transportation tool type, energy type and consumption.

[0021] Match the types of energy consumption with the preset carbon emission coefficient table. The carbon emission coefficient table stores the unit carbon emissions corresponding to different energy types. For example, the carbon emission coefficient of diesel is a fixed value of carbon dioxide generated per liter of diesel, and the carbon emission coefficient of electricity is set as a fixed value of carbon dioxide generated per kilowatt-hour of electricity according to the energy structure of the regional power grid. The matching process is to find the corresponding carbon emission coefficient according to the actual energy consumption type of the transportation tool, and multiply the consumption by the carbon emission coefficient to obtain the carbon emission intensity in the transportation stage.

[0022] Classify the data of the supply chain node according to the preset industry classification standard. The industry classification standard is divided into raw material extraction, primary processing, deep processing, and logistics transportation according to the business type of the supply chain node. The classified data is stored in the storage path corresponding to the classification label. For example, the logistics transportation data is stored in the logistics transportation database directory of the specified server, and the raw material extraction data is stored in the raw material extraction database directory.

[0023] Generate a timestamp sequence based on the regional location of the data collection terminal. The generation method of the timestamp sequence is to determine the time zone to which it belongs according to the geographical location of the data collection terminal, convert the data collection time to Coordinated Universal Time and append the regional location code. The accuracy of the timestamp sequence is set according to the data update frequency. For example, the timestamp corresponding to data updated hourly is accurate to the minute level, and the timestamp corresponding to data updated daily is accurate to the hour level. The association method between the timestamp sequence and the data update frequency is to assign high-precision timestamps to high-frequency updated data and low-precision timestamps to low-frequency updated data.

[0024] The methods for obtaining life cycle assessment data include calling the production database and logistics database of supply chain nodes through application programming interfaces. The power consumption values in the production database are collected by meter sensors and uploaded to the cloud server. The transportation records in the logistics database are uploaded through vehicle-mounted terminals or logistics management platforms. The preset industry classification standard is defined through a configuration file, and the configuration file contains the mapping relationship between classification labels and storage paths. For example, a line of data in the configuration file is "Logistics transportation: / data / logistics / ", indicating that logistics transportation data is stored in this path.

[0025] The setting basis of the carbon emission coefficient table is to obtain the unit carbon emissions of different energy types through experimental measurements or industry reports. For example, the carbon emission coefficient of diesel is determined through laboratory combustion tests, and the carbon emission coefficient of electricity is calculated by statistically analyzing the energy structure of the regional power grid and the emission data of power plants. The energy structure of the regional power grid includes the proportion of coal-fired power generation, natural gas-fired power generation, and renewable energy power generation, and the regional power grid emission factor is calculated by weighting according to this ratio.

[0026] Among them, the energy structure is obtained from the annual power industry statistical report released by the National Energy Administration, covering the installed capacity and power generation proportion of various power sources such as thermal power, hydropower, and wind power. The power plant emission data is sourced from the flue gas online monitoring system connected to the Ministry of Ecology and Environment, collecting the carbon dioxide emissions per unit of power generation of each thermal power plant. During data fusion, the carbon emission coefficients of different power generation types are weighted and calculated according to their power generation proportions. For example, the carbon emission coefficient of thermal power is multiplied by the proportion of thermal power generation, and the products of the carbon emission coefficients of other power sources and their power generation proportions are superimposed to finally form the carbon emission coefficient of electricity.

[0027] For the calculation of the regional grid emission factor, the grid carbon emission factor database released by the national or provincial ecological environment department is used as the data source, and the weights are allocated according to the proportion of the annual electricity consumption of different administrative regions within the coverage of each regional grid. In specific operations, the benchmark emission factor of each regional grid is multiplied by the proportion of electricity consumption in its corresponding administrative region and then summed to obtain the comprehensive emission factor of the regional grid. For example, a certain region contains two administrative regions, A and B. The emission factors of the grids in the two regions are weighted and averaged according to the electricity consumption ratios of 60% and 40% respectively.

[0028] The execution process of data classification is to parse the enterprise registration information or business description text of the supply chain nodes, and match the classification labels in the industry classification standard. For example, the nodes whose enterprise registration information contains the keyword "mineral mining" are classified as raw material mining category, and the nodes whose business description contains the keyword "goods transportation" are classified as logistics transportation category. The classified data is transmitted through the data pipeline to the corresponding storage path.

[0029] The generation tool for the timestamp sequence is the timestamp service deployed on the data collection terminal. The timestamp service determines the regional location according to the network protocol address or the global positioning system coordinates of the collection terminal. The regional location code adopts the area code defined by the International Organization for Standardization. For example, the regional code of Beijing, China is CN-BJ. The timestamp service converts the collection time into Coordinated Universal Time and then appends the regional code to generate a complete timestamp.

[0030] The association rule between the timestamp accuracy and the data update frequency is managed through a configuration table. The configuration table contains three columns: data source identifier, update frequency, and timestamp accuracy. For example, the data update frequency of the data source identifier "transport node A" is once per hour, and the corresponding timestamp accuracy is at the minute level. The data update frequency of the data source identifier "mining node B" is once per day, and the corresponding timestamp accuracy is at the hour level.

[0031] The regional location verification method of the data collection terminal is to reverse-lookup the geographical location through the network protocol address or directly read the global positioning system coordinates of the terminal. The reverse-lookup of the network protocol address is implemented through a third-party geolocation interface, and the global positioning system coordinates are obtained through the positioning module built into the terminal. The verified regional location is used to generate the regional code in the timestamp sequence; The update mechanism of the preset industry classification standard is to regularly receive the new classification rules uploaded by users. For example, when a user adds a new classification label "packaging material production" and associates the storage path, the system writes the new classification rules into the configuration file, and the updated classification standard will be automatically applied during subsequent data classification.

[0032] S2. Screen the incremental data updated in the last preset number of cycles according to the timestamp sequence, and predict the range of the next cycle's dynamic environmental protection constraint threshold based on the historical constraint change trajectory to determine whether the incremental data deviates. The specific implementation is as follows: Filter out the incremental data updated in the last preset number of cycles according to the chronological order of the timestamp sequence. The generation method of the timestamp sequence is to generate and append a regional code according to the regional location of the data acquisition terminal in step S1. The filtering process is to sort the timestamps from new to old, and intercept the first preset items of data as the incremental data. The number of the first preset items of data is set according to the actual business requirements. For example, if the business requirements require analyzing the data changes in the last 3 cycles, the number of the first preset items of data is set to 3. The time length of each cycle is associated with the business scenario. For example, if the environmental protection policy is adjusted quarterly, the cycle is set to 3 months.

[0033] Extract the change data of the dynamic environmental protection constraint threshold in the historical cycle. The time range of the historical cycle is consistent with the generation rule of the timestamp sequence in step S1. The change data of the dynamic environmental protection constraint threshold includes the maximum value, minimum value, and average value of the carbon emission intensity in each cycle. The time window length of the historical cycle is set according to the business requirements. For example, when analyzing the threshold change trend in the past 1 year, the time window length is set to 12 months.

[0034] Construct the threshold change trend within the time window according to the change data. The construction method is to arrange the dynamic environmental protection constraint thresholds of each cycle in chronological order, and fit the change trend line of the threshold over time through the linear regression method. The linear regression method is a data fitting method known in the prior art, and its implementation process is to calculate the slope and intercept between the time series and the threshold series to generate the trend line equation.

[0035] Calculate the range of the dynamic environmental protection constraint threshold for the next cycle through the threshold change trend. The calculation method is to substitute the time point of the next cycle into the trend line equation to obtain the predicted threshold reference value. The upper and lower limit values of the dynamic environmental protection constraint threshold range are set according to the historical threshold fluctuation range. For example, if the historical threshold fluctuation range is ±5% of the reference value, the upper limit value of the dynamic environmental protection constraint threshold range is 105% of the reference value, and the lower limit value is 95% of the reference value.

[0036] Compare the carbon emission intensity in the incremental data with the upper and lower limit values of the dynamic environmental protection constraint threshold range. The carbon emission intensity in the incremental data is sourced from the carbon emission intensity of the supply chain nodes obtained in step S1. The comparison process is to read one by one the carbon emission intensity values of each supply chain node in the incremental data. If the value is greater than the upper limit value or less than the lower limit value of the dynamic environmental protection constraint threshold range, it is determined as a deviation.

[0037] The update mechanism of the dynamic environmental protection constraint threshold is to periodically receive the threshold adjustment parameters input by the user. For example, if the user lowers the upper limit of the carbon emission intensity by 10% according to the latest environmental protection policy requirements, the system will update the adjusted threshold to the dynamic environmental protection constraint threshold range and automatically apply the new parameters in the next cycle prediction.

[0038] The adjustment rule for the time window length is dynamically adapted according to the data update frequency. For example, when the data update frequency is once a day, the time window length is set to 7 days. When the data update frequency is once a month, the time window length is set to 12 months. The association between the time window length and the data update frequency is managed through the configuration table.

[0039] The verification method for threshold change trend is to divide historical data into training set and test set. The training set is used to construct trend line, and the test set is used to verify the accuracy of dynamic environmental protection constraint threshold range. If the deviation between the actual threshold in the test set and the dynamic environmental protection constraint threshold range exceeds the preset fault tolerance rate (for example, 2%), the time window length or trend line fitting method is readjusted.

[0040] Trend line fitting method selection rules: When the monthly carbon emission data fluctuates smoothly, linear regression modeling is used; if the data presents a curve feature (such as the monthly data of Shandong Province in 2023), it will automatically switch to quadratic polynomial regression, and through the goodness of fit (R² value) comparison, the model with a higher R² value will be selected. When the data fluctuates abnormally (such as the surge in electricity consumption due to the cold wave in January 2024), the system switches to exponential smoothing, sets the weight of recent data to 0.3 and the weight of historical data to 0.7, and generates a revised trend forecast value.

[0041] The application scenarios of dynamic environmental constraint threshold ranges include environmental compliance pre-review and supply chain adjustment prediction. For example, if it is found in the pre-review stage that the carbon emission intensity of a supplier may exceed the threshold of the next cycle, the alternative supplier screening process will be initiated in advance, and nodes with high deviation risks will be identified in the prediction stage and the company will be notified to optimize the production process in advance.

[0042] The dynamic environmental constraint threshold range is visualized by integrating the dynamic environmental constraint threshold range and the historical threshold change trend into a line graph. The deviation status of the current incremental data is marked in the line graph, the deviated nodes are marked with highlighted colors, and the non-deviant nodes are displayed in standard colors. The visual chart is displayed through the front-end interface for user interactive analysis.

[0043] The storage method of the threshold comparison result is to bind the deviation judgment result and the incremental data and store them in the analysis result database. The storage path of the analysis result database is consistent with the classification storage path in step S1. For example, the deviation result of the logistics and transportation node is stored in a subfolder under the logistics and transportation database directory.

[0044] The dynamic adjustment mechanism of preset item data is to automatically expand or shrink the capacity according to the business load. For example, the number of preset items is automatically increased during business peak periods to cover more historical periods, and the number of preset items is reduced during business trough periods to reduce computing resource consumption. The trigger condition for dynamic adjustment is the real-time data processing pressure value monitored by the system.

[0045] The rule for handling abnormal data in the linear regression method is to detect and remove the outliers in the historical thresholds. The outlier detection method is to calculate the standard deviation of the threshold sequence. If the deviation of a certain threshold from the mean exceeds 3 times the standard deviation, it is determined as an outlier. After removal, the data is refitted to the trend line.

[0046] The calculation method for the confidence interval of the dynamic environmental protection constraint threshold range is to adjust the interval width according to the historical fitting error rate. For example, if the average error rate of the historical trend line is 2%, the confidence interval width is set to ±2% of the reference value. After the confidence interval is superimposed on the dynamic environmental protection constraint threshold range, the final reference threshold band is generated.

[0047] The data deduplication mechanism in the incremental data screening process is to detect and merge duplicate records of the same supply chain node in adjacent cycles. The merging rule is to retain the data with the latest timestamp, and the data with the old timestamp is marked as the historical version and archived for storage. The deduplicated incremental data is used for subsequent deviation determination.

[0048] The impact analysis of regional coding on threshold prediction is to set independent dynamic environmental protection constraint threshold ranges according to the differences in environmental protection policies in different regions. For example, nodes with the regional coding of CN - BJ apply the carbon emission standards of Beijing, and nodes with the regional coding of US - CA apply the carbon emission standards of California. The threshold ranges corresponding to the regions are managed through the regional configuration table.

[0049] The binding rule between the business scenario and the cycle length is associated through the business type label. For example, supply chain nodes with the business type of fast - moving consumer goods are set to a monthly cycle (1 month), and supply chain nodes with the business type of heavy machinery are set to a quarterly cycle (3 months). The cycle length is automatically assigned according to the business type label in the data classification stage of step S1.

[0050] The manual correction interface for the dynamic environmental protection constraint threshold range allows users to manually adjust the prediction results. For example, if a user temporarily lowers the threshold upper limit according to the internal environmental protection target (such as from 1000 tons to 900 tons), the corrected threshold range only affects the determination of the current cycle, and the prediction for the next cycle is still automatically calculated based on historical data.

[0051] The backup mechanism for historical threshold change data is to regularly package and compress the threshold data and prediction results and store them in a remote disaster - tolerant server. The backup cycle is the same as the update frequency of the timestamp sequence in step S1. For example, if the timestamp sequence is updated once an hour, the backup operation is executed once an hour.

[0052] The priority sorting rule for incremental data deviation determination is calculated by weighting according to the importance of node business. For example, the deviation determination priority of core supplier nodes is high, and that of ordinary supplier nodes is low. The determination results of high - priority nodes immediately trigger warning notifications, while those of low - priority nodes are processed with a delay.

[0053] The periodic review process of the threshold change trend is to refit the trend line every fixed period (e.g., 30 days) and compare it with the historical trend. If the change in the slope of the trend line exceeds the preset sensitivity threshold (e.g., 5%), the time window is re-divided and the prediction model is updated. The sensitivity threshold is set according to the business risk tolerance.

[0054] S3. If there is a deviation, construct a carbon emission trend vector of the time series based on the historical data of the supply chain nodes, and mark the mutation nodes exceeding the preset threshold in the carbon emission trend vector as the optimization area. The specific implementation is as follows: Extract the carbon emission intensity data of the supply chain nodes within the historical period. The time range of the historical period is the same as the length of the time window set in step S2. The source of the carbon emission intensity data is the life cycle assessment data obtained in step S1. The extraction method is to screen out the data within the corresponding time range according to the time stamp sequence and arrange them in chronological order to generate a carbon emission intensity sequence. For example, when the length of the time window is 3 months, extract the daily or weekly carbon emission intensity records of each supply chain node in the past 3 months and sort them by date.

[0055] Perform a sliding window process on the carbon emission intensity sequence. The size of the sliding window is set according to business requirements. For example, set the window size to 30 days and the sliding step to 7 days. The processing method is to divide the sequence into multiple subsequences according to the window size, and each subsequence corresponds to a time window. Calculate the mean and volatility of the carbon emission intensity within each window. The calculation method of volatility is the standard deviation or coefficient of variation of the data within the window. For example, when the carbon emission intensity within the window is 100 tons, 105 tons, and 98 tons, the mean is 101 tons, the standard deviation is 3.5 tons, and the volatility is the standard deviation divided by the mean to get 3.47%.

[0056] Combine the mean and volatility into a multi-dimensional trend vector. The dimension of the multi-dimensional trend vector is the same as the number of windows. The value of each dimension is the combination of the mean and volatility of the corresponding window. For example, when the number of windows is 12, the multi-dimensional trend vector contains 24 elements (12 means and 12 volatilities). The multi-dimensional trend vector is stored in the analysis database for subsequent steps to call.

[0057] Calculate the deviation value between the mean carbon emission intensity of each node in the multi-dimensional trend vector and the current dynamic environmental protection constraint threshold range. The current dynamic environmental protection constraint threshold range comes from the threshold range predicted in step S2. The calculation method of the deviation value is the percentage of the difference between the mean and the median of the threshold range. For example, when the threshold is from 100 tons to 110 tons, the median is 105 tons. If the mean of a certain node is 115 tons, the deviation value is (115 - 105) / 105 × 100% = 9.52%.

[0058] If the deviation value exceeds the preset fluctuation threshold, which is set according to the fluctuation characteristics of historical data. For example, when the maximum historical volatility is 10%, the preset fluctuation threshold is set to 15%. It is determined as a mutation node, and the mutation node and its associated supply chain nodes are marked as the optimization area. The determination rule for the associated supply chain nodes is to query the business association data of the supply chain nodes. The business association data includes contract binding relationships and geographical proximity relationships. For example, if the mutation node is Supplier A, and there is a contract binding relationship between Supplier A and Logistics Company B, then Logistics Company B is marked as an associated node.

[0059] The marking rule for the optimization area is to expand the marking range according to the business association of the supply chain nodes. If there is a contract binding or geographical proximity relationship between the mutation node and the associated node, the associated node is synchronously added to the optimization area. For example, if the mutation node is Supplier C located in Beijing, and the geographical proximity rule is set within the same city, then other suppliers D and E in Beijing are marked as associated nodes.

[0060] The storage method for the optimization area is to bind and store the marking results with the supply chain node information in the optimization area database. The storage path of the optimization area database is associated with the classification storage path in Step S1. For example, the optimization area marking results of logistics transportation nodes are stored in the optimization area subfolder under the logistics transportation database directory.

[0061] The dynamic adjustment mechanism for the preset fluctuation threshold is to update it periodically according to historical data. For example, the volatility data for the past year is statistically analyzed every quarter. If the volatility distribution changes, the threshold is adjusted. The adjustment method is to set the threshold to 1.5 times the historical volatility mean.

[0062] The acquisition method for business association data is to access the contract database and geographical information database of the supply chain management system. The contract database stores the cooperation information between suppliers and logistics providers, and the geographical information database stores the longitude and latitude coordinates of the nodes. The association determination is to calculate the contract intersection or geographical distance between the nodes. For example, the determination method for the contract intersection is whether both parties have signed a cooperation agreement for the same project, and the determination method for the geographical distance is to calculate whether the straight-line distance between the coordinates of the two nodes is less than 50 kilometers.

[0063] The visualization display method for the multi-dimensional trend vector is to display the mean and volatility in a double-axis line chart. The positions and deviation values of the mutation nodes are marked in the line chart. The optimization area nodes are marked in red highlight, and the non-optimization area nodes are marked in green. The visualization chart is provided for users to interactively query through the front-end interface.

[0064] The parameter optimization mechanism for the sliding window processing is to automatically adjust the window size and step length according to the data density. For example, when the data update frequency is once a day, the window size is set to 7 days and the step length is set to 1 day. When the data update frequency is once a month, the window size is set to 3 months and the step length is set to 1 month.

[0065] The exception handling rule for volatility calculation is to detect whether all data within the detection window is zero or the proportion of missing values exceeds 50%. If an anomaly is detected, the window is skipped and the window range is redetermined. For example, if a window has too many missing values due to data collection failures, the system automatically skips the window and merges the data of adjacent windows.

[0066] The priority sorting rule for associated node marking is weighted according to the business association strength. For example, the weight of the association strength bound by a contract is 0.8, and the weight of geographical proximity is 0.5. Associated nodes with a weighted value exceeding 0.6 are marked first.

[0067] The backup mechanism for the optimization area database is to regularly synchronize the marking results to a disaster recovery server in a different location. The backup period is the same as the update frequency of the timestamp sequence in step S1. For example, if the timestamp sequence is updated hourly, the optimization area data is backed up hourly.

[0068] The historical version management method for the optimization area is to store the timestamp of each marking and the judgment basis. For example, the marking time is October 5, 2023, and the judgment basis is that the deviation value exceeds 15%. Historical versions can be queried and rolled back by the timestamp.

[0069] The verification process for business association data is to regularly verify the validity of contracts and geographical information. For example, the association mark is automatically removed after the contract expires, and the proximity relationship is recalculated after the geographical information changes. The verification period is set to once a month or once a quarter according to business requirements.

[0070] The warning notification method for nodes in the optimization area is to push the marking results to the terminal devices of relevant responsible persons. The pushed content includes the node name, deviation value, and optimization suggestions. For example, the pushed message is "The carbon emission intensity deviation value of Supplier F is 18%. It is recommended to replace low-carbon materials or adjust the transportation route."

[0071] The correction interface for the deviation value allows users to manually adjust the calculation result. For example, the user adjusts the deviation value from 15% to 12% according to the actual situation. The corrected data only affects the current optimization area marking and does not modify the historical record.

[0072] The persistent storage format for the multi-dimensional trend vector is JSON or binary files. When storing, a timestamp and a data source identifier are attached. For example, the JSON file contains fields such as timestamp, node ID, mean list, and volatility list.

[0073] The parallel computing optimization method for sliding window processing is to divide large-scale data into multiple subtasks and distribute them to the computing cluster for processing. For example, a 12-month sequence is split into 4 three-month subtasks for parallel calculation of the mean and volatility.

[0074] The automatic execution trigger condition of the optimization area marking is the deviation determination result in step S2. For example, after step S2 determines that a certain node is deviated, the system automatically triggers the optimization area marking process of step S3 without manual intervention.

[0075] The dynamic update mechanism of geographic proximity is to receive node location change information in real time. For example, when a supplier moves to a new address, the system recalculates the distance between it and surrounding nodes and updates the associated tags.

[0076] The timeliness management method for contract binding relationships is to trigger an early warning a preset number of days before the contract expires, for example, notifying the user to renew or remove the association mark 30 days before the contract expires, to ensure the real-time accuracy of the optimization zone data.

[0077] S4. Generate a set of candidate solutions based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes. The specific implementation is as follows: The carbon emission intensity data of the mutation node in the optimization area and the business logic rules of the associated supply chain nodes are obtained. The business logic rules are derived from the optimization area nodes marked in step S3 and their correlation determination results. The contract binding relationship is obtained by querying the contract database of the supply chain management system. The contract database stores the cooperation agreement and validity period between the supplier and the logistics company. When extracting the upstream and downstream nodes that have a direct supply and demand relationship with the mutation node, if the mutation node is supplier A, the upstream and downstream nodes include the manufacturer B that purchases raw materials from supplier A and the logistics company C responsible for transportation, forming a supply chain carbon flow impact chain. The data structure of the supply chain carbon flow impact chain is a chain list containing node ID, carbon emission intensity and cooperation relationship.

[0078] Based on the supply chain carbon flow impact chain, the superposition effect of alternative nodes on the overall supply chain carbon emissions is analyzed. The analysis method is to assume that supplier A is replaced by low-carbon supplier D, calculate the carbon emission intensity of supplier D, and predict the incremental carbon emissions of manufacturer B due to process adjustments caused by changes in raw materials and the carbon emissions difference caused by changes in transportation routes of logistics company C. The superposition effect is the algebraic sum of the three. For example, the carbon emissions of supplier D are reduced by 50 tons, manufacturer B increases by 10 tons, and logistics company C reduces by 20 tons. The total effect is a reduction of 60 tons. When generating a set of alternative schemes, each scheme records the predicted changes in the carbon emission intensity of the alternative node D and the associated nodes B and C.

[0079] Screen node combinations for mergeable transportation routes based on geographical proximity. The geographical proximity is derived from the geographical location data of the optimized area nodes marked in step S3. The proximity range is set to 50 kilometers according to business requirements. The screening method is to find nodes with overlapping transportation paths within the same geographical range. For example, both supplier E and supplier F need to transport goods to manufacturer G, and the distance between their warehouses is 30 kilometers. Then the merged transportation route is to distribute uniformly from the central warehouse. The path priority stratification mechanism is implemented by accessing the delay sensitivity data of the production scheduling system. The delay sensitivity is divided into three levels: high, medium, and low. High-sensitivity paths are those whose delays will cause the production line to stop. Each solution in the collaborative solution set is marked with the reduced value of carbon emission intensity after route optimization and the priority label.

[0080] Perform multi-dimensional weighted fusion of the alternative solution set and the collaborative solution set based on the carbon emission intensity threshold and path priority. The carbon emission intensity threshold is derived from the predicted dynamic environmental protection constraint threshold range in step S2. The path priority weights are assigned according to the sensitivity level. For example, the high-sensitivity weight is 0.6, the medium-sensitivity weight is 0.3, and the low-sensitivity weight is 0.1. The weighted fusion formula is comprehensive score = carbon emission reduction value × 0.7 + path priority weight × 0.3. Eliminate solutions whose superposition effect exceeds the preset risk threshold. The preset risk threshold is set according to the failure rate of solutions in historical data. For example, for solutions with a failure rate exceeding 20%, the risk threshold is that the comprehensive score is lower than 60 points. Only retain solutions with scores higher than the threshold when generating the candidate solution set.

[0081] Perform a pre-check for process compatibility conflicts for each solution in the candidate solution set. The pre-check rule is to obtain the production parameter range of the alternative node. The production parameter range includes material density, temperature resistance limit, etc., and compare it with the limit value of the current process equipment. For example, if the material density of the alternative node is 2.5 g / cm³, while the maximum load-bearing density of the current equipment is 2.0 g / cm³, it is determined as a conflict. The pre-check result of the conflict triggers the solution dynamic adjustment mechanism. The adjustment mechanism is to automatically match alternative nodes with a density ≤ 2.0 g / cm³ to regenerate the solution, or suggest upgrading the equipment and recalculating.

[0082] The dynamic update mechanism of the supply chain carbon flow impact chain is to monitor the change information of the contract database in real time. For example, after the contract between supplier A and manufacturer B expires, automatically remove their upstream and downstream associations and update the impact chain data to ensure that the candidate solutions are generated based on the latest business relationships.

[0083] The exception handling rule for path priority stratification is to detect whether there is a new delay risk for high-sensitivity paths due to optimization. For example, if the single transportation volume exceeds the vehicle limit after merging the transportation routes, the system automatically splits the route and downgrades the priority label, and recalculates the reduced value of carbon emission intensity at the same time.

[0084] The weight adjustment interface for multi-dimensional weighted fusion allows users to modify the weight ratio of carbon emissions and priority according to actual needs. For example, the carbon emission weight is adjusted from 0.7 to 0.8, and the priority weight is adjusted from 0.3 to 0.2. The adjusted weights only affect the scenario generation in the current cycle.

[0085] The fault tolerance mechanism for pre-checking process compatibility conflicts is to set the parameter deviation tolerance range. For example, a density deviation within ±0.1 g / cm³ is considered compatible. If the parameter of the alternative node is 2.1 g / cm³ and the current equipment limit is 2.0 g / cm³, a warning is prompted but the node is not forced to be excluded, and it is up to the operator to confirm whether to accept the deviation.

[0086] The persistent storage method for the candidate scenario set is to package the scenario details, comprehensive score, and conflict pre-check results into a structured data file and store it in the scenario database. The storage path of the scenario database is associated with the classification storage path in step S1. For example, alternative scenarios are stored in the alternative scenario directory, and collaborative scenarios are stored in the collaborative scenario directory.

[0087] The visualization method for the supply chain carbon flow impact chain is to draw the carbon emission flow relationship between nodes as a Sankey diagram, which marks the carbon emission intensity of each node and the impact value of the alternative scenario. High-carbon emission nodes are marked in red, and low-carbon nodes are marked in green. Users can view the detailed data through the interactive interface.

[0088] The cost-benefit analysis of path merging and optimization is to calculate the ratio of the transportation cost savings value to the carbon emission reduction value after route optimization. For example, a cost savings of 100,000 yuan corresponds to a carbon emission reduction of 50 tons, and the ratio is 0.2 ten thousand yuan / ton. Scenarios with a ratio lower than the industry benchmark value are marked as high-cost performance scenarios.

[0089] The dynamic learning mechanism for the preset risk threshold is to automatically adjust the threshold according to the execution results of historical scenarios. For example, if the success rate of scenarios with a score higher than 60 points in the past 10 scenarios is 90%, the risk threshold is set to 60 points. If the success rate drops to 80%, the threshold is raised to 65 points.

[0090] The version management method for the candidate scenario set is to attach a timestamp and version number when each scenario is generated. Users can trace back to historical versions and compare the optimization effects in different cycles. For example, compare the difference in carbon emission reduction values between scenarios in September and October 2023.

[0091] The standard process for process compatibility parameters is to uniformly convert the equipment limit values to the International System of Units and store them in the parameter library. For example, convert the load-bearing density from pounds per cubic inch to g / cm³ to ensure the comparability of parameters from different data sources.

[0092] The intelligent recommendation logic for alternative node matching is to preferentially select nodes in similar scenarios based on historical successful cases. For example, in a certain solution, Supplier D successfully replaced Supplier A. Then, when Supplier E has a similar business to A, D-type nodes are preferentially recommended.

[0093] The dynamic calibration process for geographical proximity relationships is to regularly receive location update data of nodes. For example, after a certain supplier relocates, the system recalculates its distance from surrounding nodes and updates the proximity relationship markers. The calibration period is set to once a month.

[0094] The automated repair suggestions for conflict pre-check are to recommend alternative solutions or equipment adjustment plans based on the type of parameter conflict. For example, when the density exceeds the standard, a low-density material supplier is recommended; when the temperature resistance is insufficient, it is recommended to increase cooling equipment. The repair suggestions are stored in the solution remarks field for users' reference.

[0095] The real-time calculation optimization for multi-dimensional weighted fusion is to split large-scale solution data into multiple subtasks for parallel processing. For example, 1000 solutions are split into 10 groups for parallel calculation of comprehensive scores to improve the generation efficiency.

[0096] The differentiated push strategy for the candidate solution set is to customize the push content according to the user role. For example, the production department receives process compatibility details, the procurement department receives alternative node cost data, and the management level receives comprehensive scores and risk analyses.

[0097] The abnormal interruption handling for the supply chain carbon flow impact chain is to detect emergency plans when key nodes in the impact chain fail. For example, when Supplier A suddenly stops production, the backup supplier list is automatically triggered and emergency candidate solutions are regenerated.

[0098] The manual intervention interface for path priority allows users to manually adjust the priority tags. For example, a medium-sensitivity path is temporarily upgraded to high-sensitivity. The adjusted tags only affect the current solution generation.

[0099] The historical data analysis for process compatibility conflicts is to count high-frequency conflict parameters and generate a report on equipment upgrade suggestions. For example, if the density conflict of a certain equipment accounts for 80%, it is recommended to purchase equipment with a higher load-bearing capacity.

[0100] S5. Conduct production feasibility verification on the candidate solution set, and conduct backward compatibility verification between the candidate solutions and the historical version life cycle assessment data, and eliminate invalid solutions. The specific implementation is as follows: Extract the process parameters and material property data from the candidate solution set. The process parameters include the temperature range, pressure limit, and material processing accuracy of the production equipment, and the material properties include density, thermal conductivity, and tensile strength. The data is sourced from the parameter details of the alternative nodes and collaborative nodes recorded in the candidate solution set generated in step S4. The compatibility database of the current production equipment stores the technical specification data provided by the equipment manufacturer. For example, the maximum operating temperature of a certain injection molding machine is 300°C, and the maximum pressure is 150 MPa. The screening process is to compare the process parameters in the candidate solution with the limit values in the compatibility database and calculate the deviation value. For example, if the required processing temperature in the candidate solution is 320°C and the equipment limit is 300°C, the deviation value is 20°C. If the deviation value exceeds the preset threshold (e.g., 10°C), it is determined to be incompatible and excluded.

[0101] Input the screened candidate solutions into the historical version life cycle assessment model. The historical version life cycle assessment model is the calculation engine of the past versions stored in step S1, such as the versions in January, April, and July 2023. The calculation engine of each version contains the carbon emission coefficients and assessment rules at that time. The verification method is to substitute the parameters of the candidate solution into each historical version to recalculate the environmental protection score. For example, substitute the material transportation distance of alternative supplier D into the transportation carbon emission calculation formula of the January 2023 version. If the score is lower than the minimum requirement of this version (e.g., 80 points), it is determined to be unqualified. The verification covers at least three consecutive historical versions and the time span is not less than six months. For example, verify 6 versions from January to June 2023 to ensure the stability of the solution under long-term data fluctuations.

[0102] If the candidate solution passes the compatibility verification of the current production equipment but fails the environmental protection score verification of the historical version, it is determined to be a failed solution and excluded. For example, if the process parameters of a certain solution meet the equipment requirements, but due to the update of the carbon emission coefficient in the January 2023 version, the score drops from 85 points to 75 points, which is lower than the required 80 points of this version, then this solution is marked as failed. The reasons for failure include process parameter conflicts and excessive fluctuations in historical version scores. The specific types of process parameter conflicts are recorded as temperature overrun, pressure overrun, or material property mismatch. The excessive fluctuations in historical version scores are recorded as the specific version number and score difference. For example, the score difference in the version 2023-04 is -15 points.

[0103] Mark the reasons for failure of the failed solutions, and store the marking results in the failed solution database. The storage path of the failed solution database is associated with the classification storage path in step S1. For example, the failed solutions related to logistics transportation are stored in the failed solution subfolder under the logistics transportation database directory. The failure reason data includes the name of the conflicting parameter, the historical version number, and the score difference, which is for reference when marking the optimization area in subsequent step S3. For example, when the material density conflict appears frequently, it triggers the update of the marking rules in the optimization area of step S3.

[0104] The version coverage of backward compatibility verification is set according to business requirements, which include data update frequency and industry compliance requirements. For example, when the environmental protection policy is updated quarterly, the coverage is set to the versions of the last four quarters, with a time span of no less than one year. The version selection rule is to preferentially cover the versions before and after data update. For example, the versions of one month before and two months after the policy takes effect are selected to ensure the compatibility of the solution during the policy transition period.

[0105] The invocation method of the historical version life cycle assessment model is to load the corresponding calculation engine from the storage path in step S1 according to the version timestamp. For example, the calculation engine file path for version 2023-01 is / LCA_model / 2023-01 / . After loading, the calculation parameters are initialized, including the material carbon emission coefficient and transportation energy consumption formula at that time. During the verification process, the independence of the calculation logic of each version is maintained to avoid parameter pollution between versions.

[0106] The indexing mechanism of the failed solution database is to establish quick query tags according to the type of failure cause. For example, it is classified by process parameter conflict types (temperature, pressure, material) or historical version numbers. Users can filter specific types of failed solutions through tags. For example, query all solutions that failed due to temperature exceeding the standard. The indexed data is associated with the candidate solution set in step S4 to facilitate tracing the original solution generation logic.

[0107] The dynamic adjustment mechanism of the preset threshold is to regularly update the limit value in the compatibility database according to the equipment aging curve. For example, after a certain equipment has been used for three years, the maximum temperature drops to 290°C. The system automatically updates the database and recalculates the deviation value to ensure that the verification result reflects the real-time state of the equipment. The threshold adjustment record is stored in the equipment maintenance log for auditing.

[0108] The tolerance rule for environmental protection score verification is to allow a certain fluctuation range of scores in historical versions. For example, it is allowed that at most two version scores are lower than the requirement and the difference does not exceed 5 points. If the scores of a certain solution in 6 versions are lower than the requirement for 3 times or the difference exceeds 5 points, it is determined to be failed. The tolerance rule is managed through a configuration table, and the configuration table contains parameters such as the number of versions and the difference threshold.

[0109] The manual review interface for failed solutions allows users to view the automatic determination results and make manual corrections. For example, a certain solution is automatically excluded due to a score difference of -6 points, but the user confirms that the difference is within the acceptable range and then manually restores it to a valid solution. The correction record is stored in the comment field of the failed solution database.

[0110] The sandbox isolation mechanism of the historical version calculation engine runs each version of the engine in an independent process to avoid system crashes caused by code conflicts between versions. For example, the engine version of January 2023 is deployed to an independent container, and the resources are automatically released after verification. The sandbox log records abnormal events during the verification process for operation and maintenance analysis.

[0111] The data archiving rule for invalidated solutions is to transfer invalidated solutions that have exceeded the retention period to cold storage regularly. For example, keep the invalidated solutions of the most recent year available for online query, and archive earlier data to offline storage. The archiving period is synchronized with the update frequency of the timestamp sequence in step S1. For example, perform the archiving operation once a month.

[0112] The root cause analysis of process parameter conflicts is to count the high-frequency conflict parameters and generate suggestions for equipment upgrades or process optimizations. For example, if the proportion of temperature exceeding the standard in a certain batch of solutions is 70%, it is recommended to purchase high-temperature resistant equipment or optimize the cooling system. The analysis results are pushed to the production management department for decision-making reference.

[0113] The trend prediction of environmental protection score fluctuations is to fit the future fluctuation curve based on the historical version score data. For example, use linear regression to predict that the score of the next version may drop by 5%, prompting users to optimize the solution in advance. The prediction results are coordinated with the dynamic threshold prediction logic in step S2 to form a closed-loop feedback.

[0114] The triggering condition for regenerating invalidated solutions is that when the optimization area marker is updated or the threshold of the compatibility database is adjusted, the candidate solution generation process in step S4 is automatically re-executed. For example, when the temperature threshold of a certain device is increased from 300°C to 320°C, the system re-verifies whether the solutions that failed due to temperature exceeding the standard before are converted to valid.

[0115] The abnormal detection mechanism for version coverage is to monitor the integrity of historical version data. For example, if a file of a certain version of the calculation engine is missing, the version is automatically skipped and marked as data incomplete, and the administrator is notified to repair it to avoid interruption of the verification process.

[0116] The encrypted storage method for the invalidated solution database is to encrypt sensitive process parameters and score data. For example, use the AES algorithm for encrypted storage, and the key management is integrated through the enterprise security system to ensure that the data confidentiality complies with industry specifications.

[0117] The parallel computing optimization for backward compatibility verification is to distribute multiple candidate solutions to different computing nodes to verify multiple historical versions simultaneously. For example, divide 100 solutions into 10 groups for parallel processing to shorten the verification time, and the computing resource allocation strategy is dynamically adjusted according to the system load.

[0118] The result visualization report of environmental protection score verification shows the comparison between the scores of each version and the threshold value in a line chart, marking the positions and differences of the failed versions. The report is output as a PDF or an interactive web page for users to download and analyze. The visualization template is consistent with the style of the threshold prediction chart in step S2.

[0119] The retry mechanism for failed solutions allows users to resubmit the verification after optimizing the process parameters. For example, after adjusting the processing temperature within the equipment limit, step S5 is run again. The number of retries is associated with the version coverage. For example, each solution can be retried at most three times.

[0120] The version rollback function of the historical version calculation engine is that when a version error is found during the verification process, it automatically switches to the correct version for recalculation. For example, when loading the version of January 2023, if a file corruption is detected, it is automatically replaced with the backup version and continues to execute. The rollback records a log for auditing.

[0121] S6. Map the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range, and output the set of environmental protection design solutions. The specific implementation is as follows: Obtain the upper limit and lower limit of the dynamic environmental protection constraint threshold range predicted in step S2. The dynamic environmental protection constraint threshold range is derived from the prediction results based on the historical constraint change trajectory in step S2. The upper limit of the threshold range is the maximum allowable carbon emission intensity for the next period, and the lower limit is the carbon emission intensity corresponding to the lowest material cost. For example, the upper limit is that the carbon emission per unit product does not exceed 1000 kg of carbon dioxide, and the lower limit is that the carbon emission per unit product is not less than 800 kg of carbon dioxide. The generation logic of the threshold range is obtained by extrapolating the historical threshold fluctuation trend.

[0122] Calculate the normalized percentage of the comprehensive environmental protection scores of the remaining candidate solutions within the dynamic environmental protection constraint threshold range. The comprehensive environmental protection score is the scoring result of the candidate solutions that passed the verification in step S5 within the predicted threshold range in step S2. The normalization calculation method is to use the difference between the score and the lower limit as the numerator, and the difference between the upper limit and the lower limit as the denominator, and then multiply the result of dividing the numerator by the denominator by 100%. For example, if the score of a certain solution is 900 kg of carbon dioxide, the lower limit is 800 kg, and the upper limit is 1000 kg, the normalized percentage is (900 - 800) / (1000 - 800) × 100% = 50%.

[0123] Screen the candidate solutions with the normalized percentage within the preset acceptable interval. The preset acceptable interval is set from 30% to 100% according to the business risk tolerance. The business risk tolerance is defined by the policy document provided by the enterprise risk management department. For example, the high-risk tolerance corresponds to the interval of 50% to 100%, and the low-risk tolerance corresponds to the interval of 30% to 100%. The screening process is to eliminate the solutions with a normalized percentage lower than 30% and retain the candidate solutions that meet the interval.

[0124] Sort the filtered candidate solutions in descending order according to the normalized percentage to generate a set of preliminary environmental protection design solutions. The sorting rule is to preferentially select the solutions with a higher normalized percentage. If the percentages are the same, refer to the comprehensive score of the candidate solutions in step S4. For example, if the normalized percentage of solution A is 90% and that of solution B is 85%, then solution A is ranked ahead. If both are 85%, then select solution B with a higher comprehensive score.

[0125] Perform weighted correction on the set of preliminary environmental protection design solutions according to the stability score of the historical version compatibility verification in step S5. The stability score is the passing rate of the candidate solution in the historical version verification. For example, if a certain solution passes 5 times in 6 historical versions, then the stability score is 83.3% (5 / 6×100%). The weighted correction method is to add the normalized percentage and the stability score according to the weights. For example, the weight of the normalized percentage is 70% and the weight of the stability score is 30%. The final weighted value is (normalized percentage×0.7 + stability score×0.3). Re-sort in descending order of the weighted value to generate the final set of environmental protection design solutions.

[0126] The update mechanism for the dynamic environmental protection constraint threshold range is to periodically receive the latest prediction results of step S2. For example, update the threshold range once a month to ensure that the mapping process is based on the latest data. The updated threshold range covers the entire time window of the next period. For example, when the next period is 3 months, the validity period of the threshold range is set to 3 months synchronously.

[0127] The visualization display method of the normalized percentage is to present the scores of the candidate solutions and the threshold range in the form of a heat map. In the heat map, the depth of the color represents the high and low of the normalized percentage. For example, dark green represents 90%-100%, light green represents 70%-90%, yellow represents 50%-70%, and red represents less than 50%. The visualization chart is provided through the front-end interface for users to interactively analyze.

[0128] The calculation rule of the stability score is to count the consecutive passing times of the candidate solution in the historical version verification. For example, if a certain solution passes 4 times in 6 consecutive versions, then the stability score is 66.7% (4 / 6×100%). The more consecutive passing times, the higher the score. If a certain solution passes all in the latest 3 versions, then the score is increased to 100%. The calculation rule is defined through a configuration file.

[0129] The storage method of the final set of environmental protection design solutions is to package the solution details, normalized percentage, and weighted value into a structured data file and store it in the solution database. The storage path of the solution database is associated with the classification storage path in step S1. For example, high-score solutions are stored in the preferred solution directory, and low-score solutions are stored in the alternative solution directory.

[0130] The weighted correction artificial intervention interface allows users to adjust the weight ratio. For example, the normalized percentage weight is adjusted from 70% to 60%, and the stability score weight is adjusted from 30% to 40%. The adjusted weights only affect the current sorting results, and the historical data retains the original weight records.

[0131] The abnormal data processing rule for the normalized percentage is to detect whether the score exceeds the threshold range. For example, if the score of a certain solution is 700 kg of carbon dioxide, which is lower than the lower limit value of 800 kg, it is marked as an abnormal solution and a warning notice is triggered. After manual review, it is decided whether to include it in the screening. The abnormal data log is stored in the audit database for traceability.

[0132] The dynamic update mechanism for the stability score is refreshed in real time according to the verification results of step S5. For example, when the verification of the new historical version is completed, the stability scores of all candidate solutions are automatically recalculated and the sorting is updated to ensure that the final set reflects the latest verification status.

[0133] The differential output strategy for the final set of environmental protection design solutions is to generate customized reports according to the user role. For example, the management level receives a concise report containing the top 10 solutions, and the technical department receives a full version report containing all solutions and technical parameters. The report format supports PDF, Excel, or interactive web pages.

[0134] The calibration mechanism for the dynamic environmental protection constraint threshold range is to adjust the prediction model for the next cycle according to the data deviation within the actual cycle. For example, when the deviation between the actual threshold range and the predicted value in this cycle exceeds 10%, the historical data window length is automatically increased or the trend fitting parameters are adjusted. The calibration records are stored in the model log for optimization reference.

[0135] The triggering condition for the remapping of candidate solutions is that when the threshold range is updated or the stability score changes exceed the preset sensitivity, the mapping process in step S6 is automatically re-executed. For example, when the threshold range changes by more than 5% or the stability score fluctuates by more than 10%, the remapping is triggered and the user is notified.

[0136] The tolerance rule for the normalized percentage screening allows some solutions to break through the interval limit under specific conditions. For example, if the normalized percentage of a certain solution is 25%, but its stability score is 95%, it can be temporarily included in the preferred catalog through manual approval. The tolerance rule is managed through the approval process configuration table.

[0137] The version management method for the final set of environmental protection design solutions is to append a timestamp and a version number each time it is generated. For example, the solution set of version 2023-10. Users can trace back to historical versions and compare the optimization effects of different cycles. The version data is regularly archived to an independent storage server to release online resources.

[0138] The parallel computing optimization of weighted sorting is to divide large-scale scenario data into multiple subsets for parallel processing. For example, 1000 scenarios are divided into 10 groups for parallel calculation of weighted values, and a distributed computing cluster is used to shorten the processing time. The resource allocation strategy is dynamically adjusted according to the system load.

[0139] The trend analysis of stability scores is to predict future stability changes based on historical score data. For example, when the stability score of a certain scenario continues to decline, a risk is prompted. The prediction results cooperate with the threshold prediction in step S2 to generate optimization suggestions, forming a closed-loop feedback.

[0140] The encryption transmission method for the final scenario set is to encrypt sensitive data using the SSL protocol and then transmit it to the user terminal to ensure that the data is not intercepted or tampered with during transmission. The encryption key is rotated regularly through the enterprise key management system.

[0141] The visualization comparison function of the dynamic environmental protection constraint threshold range is to display the predicted threshold and the historical threshold in a superimposed line chart, mark the distribution position of the current scenario score, and users can drag the time axis to view the threshold change trend in different periods.

[0142] The automated report generation of normalized percentage screening is to automatically generate a summary and analysis conclusion based on the screening results. For example, "The proportion of preferred scenarios is 60%, and the average normalized percentage is 85%". The report template supports custom fields and chart types.

[0143] Embodiment 2: Figure 2 The structural schematic diagram of an environmental protection design scheme generation system of the present invention is given. An environmental protection design scheme generation system includes the following modules: The data acquisition module is used to acquire the life cycle assessment data of the target product and its time stamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain nodes; The dynamic prediction module is used to screen the incremental data updated in the last preset number of cycles according to the time stamp sequence, and predict the range of the dynamic environmental protection constraint threshold in the next cycle based on the historical constraint change trajectory to determine whether the incremental data deviates; If it deviates, the optimization marking module constructs a carbon emission trend vector of the time series according to the historical data of the supply chain nodes, and marks the mutation nodes exceeding the preset threshold in the carbon emission trend vector as the optimization area; The scheme generation module is used to generate a candidate scheme set based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes; The verification and elimination module is used to verify the production feasibility of the candidate scheme set, and perform reverse compatibility verification on the candidate scheme and the life cycle assessment data of the historical version, and eliminate invalid schemes; A result output module, configured to map the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range, and output a set of environmental protection design solutions.

[0144] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula that is closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0145] It should be noted that the present invention can be deployed on the device itself to achieve embedded applications, or can also run on a PC or other terminal with a user interface, so as to meet various hardware environments and usage requirements.

[0146] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product.

[0147] Those of ordinary skill in the art can realize that the modules and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application of the technical solution and the invention constraints. Those skilled in the art can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0148] In addition, in each embodiment of the present application, the functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0149] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces, and the indirect coupling or communication connection of the device or module can be in an electrical, mechanical or other form.

[0150] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the said claims.

[0151] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for generating an environmental design scheme, characterized in that: The steps include: S1. Obtain the life cycle assessment data of the target product and its timestamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain nodes. S2. Filter the incremental data updated for the last preset number of times according to the timestamp sequence, and predict the dynamic environmental protection constraint threshold range of the next period based on the historical constraint change trajectory to determine whether the incremental data deviates; S3. If there is a deviation, a carbon emission trend vector of the time series is constructed based on the historical data of the supply chain nodes, and the mutation nodes in the carbon emission trend vector that exceed the preset threshold are marked as optimization areas; S4. Generate a set of candidate solutions based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes; S5. Verify the production feasibility of the candidate solution set, and verify the reverse compatibility of the candidate solution with the historical version life cycle assessment data to eliminate invalid solutions; S6. Map the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range, and output a set of environmental protection design solutions.

2. The method for generating an environmental design scheme according to claim 1, characterized in that: Obtain the life cycle assessment data of the target product and its timestamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain nodes, including: Obtain the carbon emissions corresponding to the raw material production stage and the energy consumption type corresponding to the transportation stage of the supply chain node, match the energy consumption type with the preset carbon emission coefficient table, and obtain the carbon emission intensity of the transportation stage; Classify data of supply chain nodes according to preset industry classification standards, and bind the classified data to the corresponding storage path; A timestamp sequence is generated according to the regional location of the data collection terminal, wherein the accuracy of the timestamp sequence is associated with the data update frequency.

3. The method for generating an environmental design scheme according to claim 1, characterized in that: Filter the incremental data updated in the last preset number of cycles according to the timestamp sequence, and predict the dynamic environmental protection constraint threshold range of the next cycle based on the historical constraint change trajectory to determine whether the incremental data deviates, including: Filter out the incremental data updated at the last preset number of times according to the time sequence of the timestamp sequence, where the incremental data updated at the last preset number of times is the previous preset item data intercepted after sorting the timestamps from new to old; Extract the change data of dynamic environmental protection constraint thresholds in historical cycles, build the threshold change trend within the time window based on the change data, and calculate the dynamic environmental protection constraint threshold range of the next cycle through the threshold change trend; Compare the carbon emission intensity in the incremental data with the upper and lower limits of the dynamic environmental constraint threshold range. If it exceeds the upper limit or is lower than the lower limit, it is judged as a deviation.

4. The method for generating an environmental design scheme according to claim 1, characterized in that: If there is a deviation, a time series carbon emission trend vector is constructed based on the historical data of the supply chain nodes, and the mutation nodes in the carbon emission trend vector that exceed the preset threshold are marked as optimization areas, including: Extract the carbon emission intensity data of supply chain nodes in the historical period, arrange them in chronological order to generate a carbon emission intensity sequence, perform sliding window processing on the carbon emission intensity sequence, calculate the mean and volatility of carbon emission intensity in each window, and combine the mean and volatility into a multidimensional trend vector; Calculate the deviation between the mean carbon emission intensity of each node in the multidimensional trend vector and the current dynamic environmental protection constraint threshold range. If the deviation exceeds the preset fluctuation threshold, it is determined to be a mutation node, and the mutation node and its associated supply chain nodes are marked as optimization areas. The marking rule of the optimization zone is to expand the marking range according to the business relevance of the supply chain nodes. If the mutation node and the associated node have a contractual binding or geographical proximity relationship, the associated node will be added to the optimization zone synchronously.

5. The method for generating an environmental design scheme according to claim 1, characterized in that: Based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes, a set of candidate solutions is generated, including: Obtain the carbon emission intensity data of the mutation nodes in the optimization area and the business logic rules of the associated supply chain nodes, extract the upstream and downstream nodes that have direct supply and demand relationships with the mutation nodes through contract binding relationships, and form a supply chain carbon flow impact chain; Based on the supply chain carbon flow impact chain, the superposition effect of alternative nodes on the overall supply chain carbon emissions is analyzed to generate a set of alternative solutions. Each alternative solution contains the predicted value of the carbon emission intensity change of the alternative node and the corresponding upstream and downstream nodes. Filter node combinations that can merge transportation routes based on geographic proximity, and generate a set of collaborative solutions based on the path priority stratification mechanism. The priority stratification mechanism is to prioritize and optimize transportation routes with high delay sensitivity. The alternative solution set and the synergistic solution set are weighted and integrated in multiple dimensions based on the carbon emission intensity threshold and path priority, and the solutions whose superposition effect exceeds the preset risk threshold are eliminated to generate a candidate solution set; A process compatibility conflict pre-check is performed on each solution in the candidate solution set.

6. The method for generating an environmental design scheme according to claim 5, characterized in that: The pre-check rule is to match the production parameter range of the alternative node with the limit value of the current process equipment, and the conflict pre-check result triggers the dynamic adjustment mechanism of the plan.

7. The method for generating an environmental design scheme according to claim 1, characterized in that: Verify the production feasibility of the candidate solution set, and verify the reverse compatibility of the candidate solution with the historical version life cycle assessment data to eliminate invalid solutions, including: Extract process parameters and material property data from the candidate solution set, compare them with the compatibility database of the current production equipment, and select solutions with process parameter deviation values ​​lower than the preset threshold; The selected candidate solutions are input into the historical version life cycle assessment model to verify whether their environmental protection scores in the historical version data continue to meet the standards; If a candidate solution passes the current production equipment compatibility verification but fails to meet the environmental score verification of the historical version, it will be judged as an invalid solution and eliminated; The failure reasons of the failed solutions are marked, including process parameter conflicts and excessive fluctuations in historical version scores. The marking results are stored in the failure solution database for subsequent optimization reference.

8. The method for generating an environmental design scheme according to claim 7, characterized in that: The method for verifying whether the environmental protection score in the historical version data continues to meet the standards is to substitute the program parameters into the calculation engine of the corresponding historical version and re-score.

9. The method for generating an environmental design scheme according to claim 1, characterized in that: The comprehensive environmental protection scores of the remaining candidate solutions are mapped to the dynamic environmental protection constraint threshold range, and a set of environmental protection design solutions is output, including: Obtain the predicted upper and lower limits of the dynamic environmental protection constraint threshold range for the next cycle; Calculate the normalized percentage of the comprehensive environmental protection scores of the remaining candidate solutions within the dynamic environmental protection constraint threshold range; Screen candidate solutions whose normalized percentages are within a preset acceptable range; Sort the selected candidate solutions from high to low according to the normalized percentage to generate a preliminary set of environmental protection design solutions; The preliminary set of environmental design solutions is weightedly revised according to the stability score of the historical version compatibility verification. The higher the stability score, the greater the sorting weight, and the final set of environmental design solutions is generated.

10. An environmental design scheme generation system, used to implement an environmental design scheme generation method according to any one of claims 1 to 9, characterized in that: Includes the following modules: A data acquisition module is used to obtain the life cycle assessment data of the target product and its timestamp sequence. The life cycle assessment data includes the carbon emission intensity of the supply chain nodes. A dynamic prediction module is used to filter the incremental data updated for the last preset number of periods according to the timestamp sequence, and predict the dynamic environmental protection constraint threshold range of the next period based on the historical constraint change trajectory to determine whether the incremental data deviates; If there is a deviation, the optimization marking module constructs a time series carbon emission trend vector based on the historical data of the supply chain nodes, and marks the mutation nodes in the carbon emission trend vector that exceed the preset threshold as the optimization area; A solution generation module is used to generate a set of candidate solutions based on the carbon emission intensity of the optimization area and the business logic association rules of the mutation nodes; The verification and elimination module is used to verify the production feasibility of the candidate solution set, and to verify the reverse compatibility of the candidate solution with the historical version life cycle assessment data to eliminate invalid solutions; The result output module is used to map the comprehensive environmental protection scores of the remaining candidate solutions to the dynamic environmental protection constraint threshold range and output a set of environmental protection design solutions.

Images