A method for constructing a supply chain optimization model for port logistics and city industry linkage

Abstract

The application discloses a kind of port logistics and city industry linkage's supply chain optimization model construction method, it is related to supply chain collaborative optimization technical field, including, collection port logistics data, city industry data and city consumption data, carry out cleaning and fusion processing, form port logistics and city industry supply chain data lake;Based on city consumption data in port logistics and city industry supply chain data lake, consumption demand is forecasted and is reversely converted into supply chain demand instruction set to industry production and port operation;According to supply chain demand instruction set, real-time data in port logistics and city industry supply chain data lake, calculate and monitor the performance index of port logistics and city industry linkage.This application realizes model iterative optimization by executing feedback, forms data-driven, causal perception and continuously evolving closed-loop collaborative decision, effectively improves the accuracy, dynamics and overall efficiency of port and industry linkage.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of supply chain collaborative optimization, and in particular to a supply chain optimization model construction method for port logistics and city industry linkage. BACKGROUND

[0002] Under the dual driving of globalization economy and regional integration development, modern supply chain management has shifted from single-link optimization to multi-system collaborative decision-making. As an international logistics hub, the operation efficiency of a port directly affects the supply of raw materials and the circulation of finished products for the industries in the hinterland. Meanwhile, the dynamic demand of city industries also shapes the operation rhythm and resource allocation strategy of the port. In order to improve the competitiveness of the regional economy, the academia and industry gradually pay attention to the linkage between the port and the industry, and try to build a collaborative model through digital means. In the prior art, a representative scheme is a supply chain collaborative framework based on a multi-agent system. In this kind of framework, the port operation subjects and the industry subjects are modeled as autonomous agents, and through defining negotiation rules and objective functions, a contract net protocol or an auction mechanism is used to realize distributed task allocation and resource scheduling.

[0003] However, although the multi-agent system has advantages in heterogeneous subject collaboration, it still has limitations in dealing with the high dynamicity and deep coupling relationship of the port-industry complex system. The existing MAS framework usually relies on pre-defined rules and local objectives for interaction, lacks global perception and dynamic optimization capability for the overall state of the system, and now it is difficult to quantify the causal relationship strength between the port operation indicators and the industry operation indicators, resulting in collaborative decision-making based on correlation rather than causality, which cannot accurately identify and prioritize the key bottleneck links. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a supply chain optimization model construction method for port logistics and city industry linkage, which solves the problem that the prior art cannot focus on key contradictions for deep collaborative optimization due to relying on correlation decision-making.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a supply chain optimization model construction method for port logistics and city industry linkage, which comprises collecting port logistics data, city industry data and city consumption data, performing cleaning and fusion processing to form a port logistics and city industry supply chain data lake;

[0008] Based on the city consumption data in the port logistics and city industry supply chain data lake, the consumption demand is predicted and inversely transformed into a supply chain demand instruction set for industry production and port operation.

[0009] According to the real-time data in the port logistics and urban industry supply chain data lake, the performance indicators of the port logistics and urban industry linkage are calculated and monitored;

[0010] Based on the performance indicators, the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage are diagnosed by using the causal inference analysis technology;

[0011] The supply chain demand instruction set is taken as the optimization target, the resource data in the port logistics and urban industry supply chain data lake is taken as the constraint, and the bottleneck link and root cause are taken as the optimization focus. A multi-objective optimization model is constructed, the multi-objective optimization model is solved, and a supply chain optimization execution scheme is generated;

[0012] The supply chain optimization execution scheme is executed, and the execution effect data is fed back to the port logistics and urban industry supply chain data lake, and the multi-objective optimization model is iteratively updated.

[0013] As a preferred scheme of the port logistics and urban industry linkage supply chain optimization model construction method, the port logistics data, the urban industry data and the urban consumption data are collected, cleaned and fused to form a port logistics and urban industry supply chain data lake, which includes the following steps,

[0014] The port logistics data, the urban industry data and the urban consumption data are obtained by acquiring the port logistics data, the urban industry data and the urban consumption data through sensing devices and data interfaces, and performing de-duplication, error correction and abnormal value processing on the port logistics data, the urban industry data and the urban consumption data;

[0015] The cleaned port logistics data, the urban industry data and the urban consumption data are converted into a unified space-time format and measurement unit to obtain standardized port logistics data, urban industry data and urban consumption data;

[0016] The standardized port logistics data, the urban industry data and the urban consumption data are aligned and associated according to entities and time to obtain fused port logistics data, urban industry data and urban consumption data;

[0017] The fused port logistics data, the urban industry data and the urban consumption data are persistently stored to form a port logistics and urban industry supply chain data lake.

[0018] As a preferred scheme of the port logistics and urban industry linkage supply chain optimization model construction method, based on the urban consumption data in the port logistics and urban industry supply chain data lake, the consumption demand is predicted and inversely transformed into a supply chain demand instruction set for industrial production and port operation, which includes the following steps,

[0019] extracting city consumption data in a one-year time range from the port logistics and city industry supply chain data lake;

[0020] Based on the city consumption data, using the association rule mining method to construct the derived features, using the long short-term memory network model to train the derived features, and generating the city consumption demand prediction report;

[0021] Through collecting the maximum operating rate of port equipment, the saturation capacity of the yard, the upper limit of road traffic capacity, the maximum load of enterprise production line and the scale of transport vehicle fleet data, and after cleaning and standardization processing, form the capacity and transport capacity constraints in the port logistics and city industry supply chain data lake;

[0022] Taking the city consumption demand prediction report as the optimization target, combining the capacity and transport capacity constraints in the port logistics and city industry supply chain data lake, and solving the industrial production instruction set and the port operation instruction set through mathematical programming;

[0023] The industrial production instruction set and the port operation instruction set are combined to form the supply chain demand instruction set.

[0024] As a preferred scheme of the port logistics and city industry supply chain optimization model construction method, according to the supply chain demand instruction set, the real-time data in the port logistics and city industry supply chain data lake, the performance indicators of the port logistics and city industry linkage are calculated and monitored, including the following steps,

[0025] Extracting port logistics real-time data and city industry real-time data from the port logistics and city industry supply chain data lake, based on historical operation cycle data, real-time ship AIS forecast data, enterprise production rhythm data and road traffic prediction data in the port logistics and city industry supply chain data lake, the planning timeline in the supply chain demand instruction set is obtained through operational research;

[0026] The planning timeline in the supply chain demand instruction set is accurately aligned with the time stamp of the port logistics real-time data and the city industry real-time data to form the plan-actual comparison data set;

[0027] Based on the plan-actual comparison data set, using time series pattern recognition and association rule mining technology to extract port indicators, industry indicators and linkage indicators, and combining the port indicators, industry indicators and linkage indicators to form the linkage performance index set;

[0028] The entropy weight method is used to assign weights to the indicators in the linkage performance index set, and the port-industry linkage comprehensive index is calculated;

[0029] By analyzing the historical linkage performance index set data in the port logistics and city industry supply chain data lake, the linkage threshold is set.

[0030] The linkage performance index set and the port production linkage comprehensive index are monitored, and when the value of the linkage comprehensive index exceeds the linkage threshold value, a warning signal is triggered.

[0031] As a preferred scheme of the port logistics and urban industry linkage supply chain optimization model construction method, based on the performance index, the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage is diagnosed by using the causal inference analysis technology, including the following steps,

[0032] The port ship average time in port and enterprise raw material inventory days indicators in the linkage performance index set and the port production linkage comprehensive index are subjected to convergence cross mapping analysis, the causal strength and direction between the port ship average time in port and enterprise raw material inventory days indicators are quantified, and the causal contribution degree between the port ship average time in port and enterprise raw material inventory days is calculated by using the convergence cross mapping algorithm;

[0033] Based on the causal contribution degree, the port ship average time in port, the average speed of the port road and the industry raw material inventory turnover days indicators in the linkage performance index set which affect the port production linkage comprehensive index are identified, and the bottleneck link is formed;

[0034] The bottleneck link is associated with the plan-actual comparison data set for analysis, the specific operation event of the bottleneck link is located, and the root cause analysis result is obtained;

[0035] The bottleneck link, the causal strength and the root cause analysis result are integrated to form the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage.

[0036] As a preferred scheme of the port logistics and urban industry linkage supply chain optimization model construction method, the supply chain demand instruction set is taken as the optimization target, the resource data in the port logistics and urban industry supply chain data lake is taken as the constraint, and the bottleneck link and the root cause are taken as the optimization focus, and a multi-objective optimization model is constructed, including the following steps,

[0037] Based on the supply chain demand instruction set and the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage, the resource data is extracted from the port logistics and urban industry supply chain data lake;

[0038] By taking the supply chain demand instruction set as the target and the resource data as the constraint, the bottleneck mathematical tightening in the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage is converted into a strengthened constraint, and a multi-objective optimization model is constructed.

[0039] As a preferred scheme of the port logistics and city industry linkage supply chain optimization model construction method, the multi-objective optimization model is solved to generate a supply chain optimization execution scheme, including the following steps,

[0040] The elite strategy is obtained by non-dominated sorting between the population in the current evolution generation and the elite individual set generated in the last generation.

[0041] The multi-objective optimization model is solved by using the fast non-dominated sorting genetic algorithm with the elite strategy to obtain a Pareto optimal solution set.

[0042] The final scheme is selected from the Pareto optimal solution set based on the cost, time efficiency, and bottleneck relief degree preference weight using the criteria decision analysis method to form the supply chain optimization execution scheme.

[0043] 9. As a preferred scheme of the port logistics and city industry linkage supply chain optimization model construction method, the supply chain optimization execution scheme is executed, and the execution effect data is fed back to the port logistics and city industry supply chain data lake for iterative updating of the multi-objective optimization model, including the following steps,

[0044] The supply chain optimization execution scheme is issued to the port operation terminal, transportation scheduling terminal, and enterprise production terminal to drive actual operation, and new port logistics data and city industry data generated by actual operation of the supply chain optimization execution scheme are collected.

[0045] The port logistics new data and city industry new data are transported back to the port logistics and city industry supply chain data lake for storage.

[0046] Based on the historical data and new data in the port logistics and city industry supply chain data lake, the parameters and weights of the multi-objective optimization model are recalibrated.

[0047] A new generation of supply chain optimization execution scheme is generated using the updated multi-objective optimization model to start a new round of optimization cycle.

[0048] In a second aspect, the present application provides a computer device comprising a memory and a processor, the memory storing a computer program, wherein the computer program is executed by the processor to implement any step of the port logistics and city industry linkage supply chain optimization model construction method according to the first aspect of the present application.

[0049] In a third aspect, the present application provides a computer readable storage medium storing a computer program, wherein the computer program is executed by the processor to implement any step of the port logistics and city industry linkage supply chain optimization model construction method according to the first aspect of the present application.

[0050] The application has the beneficial effects that: a supply chain data lake is constructed by collecting and fusing multi-source data, demand instruction sets are generated based on consumption data prediction, linkage performance indexes are calculated and monitored in real time, bottleneck links and roots are accurately diagnosed by using causal inference technology, a multi-objective optimization model is constructed and solved to generate an execution scheme by taking the instruction sets as targets, resource data as constraints, and bottlenecks as focuses, model iteration optimization is realized through execution feedback, a closed-loop collaborative decision is formed which is data-driven, causally-aware and continuously evolving, and the precision, dynamics and overall efficiency of port and industry linkage are effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0052] Fig. 1 It is a flow chart of a port logistics and city industry linkage supply chain optimization model construction method.

[0053] Fig. 2 It is a schematic diagram of a port logistics and city industry supply chain data lake.

[0054] Fig. 3 It is a schematic diagram of triggering an early warning signal.

[0055] Fig. 4 It is a schematic diagram of a supply chain demand instruction set. DETAILED DESCRIPTION

[0056] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification.

[0057] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0058] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0059] REFERENCE Figs. 1-4For an embodiment of the present application, the embodiment provides a port logistics and urban industry linkage supply chain optimization model construction method, comprising the following steps:

[0060] S1, collect port logistics data, urban industry data and urban consumption data, and perform cleaning and fusion processing to form a port logistics and urban industry supply chain data lake.

[0061] S1.1, obtain port logistics data, urban industry data and urban consumption data through sensing devices and data interfaces, and perform deduplication, error correction and outlier processing on the port logistics data, urban industry data and urban consumption data to obtain cleaned port logistics data, urban industry data and urban consumption data.

[0062] Further, through the deployment of Internet of Things sensing devices in port cranes, yards, trucks and warehouses, real-time capture of device status, container displacement and storage information is realized, and through application programming interfaces, enterprise resource planning systems, warehouse management units, e-commerce platform databases and sales terminals are connected to obtain production plans, inventory levels, sales records and consumption trend information. The obtained port logistics data, urban industry data and urban consumption data are subjected to redundant record elimination, logical error correction and abnormal value filtering beyond a reasonable range to finally obtain cleaned port logistics data, urban industry data and urban consumption data.

[0063] S1.2, convert the cleaned port logistics data, urban industry data and urban consumption data into a unified spatiotemporal format and measurement unit to obtain standardized port logistics data, urban industry data and urban consumption data.

[0064] Further, the time stamps in the cleaned port logistics data, urban industry data and urban consumption data are uniformly converted into Coordinated Universal Time format, geographic coordinates are mapped to a standard latitude-longitude coordinate system, and cargo weight, volume, amount and other measurement units are uniformly converted into international standard units or industry common units to obtain standardized port logistics data, urban industry data and urban consumption data.

[0065] S1.3, align and associate the standardized port logistics data, urban industry data and urban consumption data by entity and time to obtain fused port logistics data, urban industry data and urban consumption data.

[0066] Further, based on entity unique identifiers such as container numbers, order numbers and vehicle identifiers, the standardized port logistics data, urban industry data and urban consumption data are time-aligned at the same time granularity, and data links between port operation events and industry production activities and consumption orders are established through association rules to obtain fused port logistics data, urban industry data and urban consumption data.

[0067] S1.4, persistently store the fused port logistics data, city industry data and city consumption data to form a port logistics and city industry supply chain data lake.

[0068] Further, the fused port logistics data, city industry data and city consumption data are persistently stored by topic partitioning using a distributed columnar database, and the topic partitioning includes port operation partitioning, industry inventory partitioning, consumption transaction partitioning, etc., to form a port logistics and city industry supply chain data lake.

[0069] S2, based on the city consumption data in the port logistics and city industry supply chain data lake, consumption demand prediction is performed and is inversely transformed into a supply chain demand instruction set for industry production and port operation.

[0070] S2.1, extract city consumption data in a one-year time range from the port logistics and city industry supply chain data lake.

[0071] Further, complete extraction of city consumption data in a specified time range by using a structured query language to filter and extract complete historical records in the city consumption data partition of the port logistics and city industry supply chain data lake in the last year, which contain commodity categories, sales quantities, transaction times and geographic location information.

[0072] S2.2, based on the city consumption data, use an association rule mining method to construct derived features, and use a long short-term memory network model to train the derived features to generate a city consumption demand prediction report.

[0073] Further, based on the extracted city consumption data, an Apriori algorithm is used for association rule mining to construct derived features such as weekly sales moving average and regional demand distribution coefficient, and then a long short-term memory network model is used to supervise training of the city consumption data containing the derived features to generate a city consumption demand prediction report containing future multi-period product category and regional demand prediction values.

[0074] S2.3, by collecting port equipment maximum operation rate, yard saturation capacity, road traffic capacity upper limit, enterprise production line maximum load and transportation fleet size data, and after cleaning and standardization processing, form the capacity and transportation capacity constraints in the port logistics and city industry supply chain data lake.

[0075] Further, by collecting the physical upper limit data such as the rated work efficiency of the port quay crane, the designed maximum capacity of the yard, the designed traffic capacity of the port access road, the theoretical maximum output of the enterprise production line, and the total number of recorded vehicles of the transport vehicle fleet, and after the same processing procedure as data cleaning and standardization, the data are persistently stored in the resource constraint partition of the port logistics and urban industrial supply chain data lake, forming the capacity and transport capacity constraints in the port logistics and urban industrial supply chain data lake.

[0076] S2.4, the urban consumption demand prediction report is taken as an optimization target, the capacity and transport capacity constraints in the port logistics and urban industrial supply chain data lake are combined, and the industrial production instruction set and the port operation instruction set are obtained by mathematical programming solution.

[0077] The expression of the industrial production instruction set is:

[0078]

[0079] wherein Pr is the industrial production instruction set, p is the product category, f is the unique identification code of the production line, and t is the production time period, is the planned quantity of the specified product produced in the specified factory at the specified time, P is the product category set, F is the unique identification code set of the production line, and T is the production time period set.

[0080] The expression of the subset of loading and unloading instructions is:

[0081]

[0082] wherein Ps is the subset of loading and unloading instructions, b is the unique identification code of the loading and unloading berth, is the planned quantity of the specified goods loaded and unloaded at the specified berth at the specified time, and B is the set of loading and unloading berths.

[0083] The expression of the subset of distribution instructions is:

[0084]

[0085] wherein Pd is the subset of distribution instructions, r is the transport path, is the planned quantity of the specified goods transported at the specified path at the specified time.

[0086] The expression of the port operation instruction set is:

[0087] Pe=Pd∪Ps;

[0088] wherein Pe is the port operation instruction set.

[0089] Further, the predicted demand in the city consumption demand prediction report is taken as the optimization target of the mathematical programming model, the production capacity and transport capacity in the port logistics and industrial supply chain data lake are taken as the constraint conditions of the model, the linear programming solver is used to solve the model, and the industrial production instruction set and the port operation instruction set are output; the industrial production instruction set specifically specifies the planned production quantity of different product categories in different production lines at a specific production time period; the port operation instruction set is composed of a loading and unloading instruction subset and a distribution instruction subset, the loading and unloading instruction subset specifies the planned loading and unloading quantity of different cargos at a specific berth at a specific time window, and the distribution instruction subset specifies the planned transportation quantity of different cargos via a specific path at a specific time window.

[0090] S2.5, the industrial production instruction set and the port operation instruction set are combined to form a supply chain demand instruction set.

[0091] Further, the industrial production instruction set and the port operation instruction set are combined to form a supply chain demand instruction set; the industrial production instruction set contains the unique identification code of the product category and the production line, the production time period, and the planned quantity of the specified product produced at the specified time in the specified factory; the port operation instruction set is composed of a loading and unloading instruction subset and a distribution instruction subset, the loading and unloading instruction subset contains the unique identification code of the loading and unloading operation berth and the quantity of the specified cargo planned to be loaded and unloaded at the specified berth at the specified time, and the distribution instruction subset contains the transportation path and the quantity of the specified cargo planned to be transported via the specified path at the specified time; the merging process aligns and integrates the industrial production instruction set and the port operation instruction set according to the unified time sequence and entity identification through data splicing technology to generate the supply chain demand instruction set.

[0092] S3, according to the supply chain demand instruction set and the real-time data in the port logistics and city industrial supply chain data lake, the performance indicators of the port logistics and city industrial linkage are calculated and monitored.

[0093] S3.1, the port logistics real-time data and the city industrial real-time data are extracted from the port logistics and city industrial supply chain data lake, the planned time line in the supply chain demand instruction set is obtained through operational research based on the historical operation cycle data in the port logistics and city industrial supply chain data lake, real-time ship AIS prediction data, enterprise production rhythm data, and road traffic prediction data.

[0094] Further, the latest cycle of port logistics real-time data and city industry real-time data are extracted from the real-time data partition of the port logistics and city industry supply chain data lake, while the historical operation cycle data are obtained from the historical data partition, the real-time ship AIS forecast data and road traffic prediction data are received from the external data interface, and the production rhythm data are collected from the enterprise production data stream. Based on these multi-source data, the plan start time and plan end time of each operation activity in the supply chain demand instruction set are planned using the critical path method in operations research, and a plan timeline of the supply chain demand instruction set is formed.

[0095] S3.2, the plan timeline in the supply chain demand instruction set is accurately aligned with the timestamps of the port logistics real-time data and the city industry real-time data, and a plan-actual comparison data set is formed.

[0096] Further, the plan timeline of each operation in the supply chain demand instruction set is aligned and matched with the actual operation start time, actual operation end time recorded in the port logistics real-time data, and the actual material arrival time, actual production start time recorded in the city industry real-time data with millisecond-level precision, and the corresponding relationship between the plan value and the actual value is established through time series alignment algorithm, and a plan-actual comparison data set is formed.

[0097] S3.3, based on the plan-actual comparison data set, the port end indicators, industry end indicators and linkage end indicators are extracted using time series pattern recognition and association rule mining technology, and the port end indicators, industry end indicators and linkage end indicators are summarized to form a generated linkage performance indicator set.

[0098] Further, based on the plan-actual comparison data set, the port ship average time in port, shore crane average operation efficiency and other port end indicators are extracted using time series decomposition algorithm, the raw material inventory turnover days, production plan completion rate and other industry end indicators are extracted using association rule mining algorithm, and the port-production linkage responsiveness, transport resource matching degree and other linkage end indicators are extracted using correlation analysis method. All extracted indicator data are collected and stored in categories to generate a structured linkage performance indicator set.

[0099] S3.4, the entropy weight method is used to assign weights to the indicators in the linkage performance indicator set, and the port-production linkage comprehensive index is calculated.

[0100] The port-production linkage comprehensive index expression is:

[0101]

[0102] Wherein, G is the port-production linkage comprehensive index, w j is the weight of the jth indicator, r jis the standardized value of the jth indicator in the current evaluation period, j is the indicator index, and n is the total number of indicators.

[0103] Further, the entropy weight method is used to calculate the entropy value of each indicator in the linkage performance indicator set, and a difference coefficient is calculated according to the entropy value. Each indicator is assigned a weight coefficient according to the size of the difference coefficient. The standardized value of each indicator is linearly weighted and summed with the corresponding weight coefficient to calculate a port-industry linkage comprehensive index that comprehensively reflects the port-industry linkage level.

[0104] S3.5, set the linkage threshold by analyzing the historical linkage performance indicator set data in the port logistics and urban industry supply chain data lake.

[0105] Further, by statistically analyzing the historical linkage performance indicator set data stored in the port logistics and urban industry supply chain data lake, the historical average value and standard deviation of the port-industry linkage comprehensive index are obtained, and the upper and lower limits of the linkage threshold are set as the historical average value plus or minus twice the standard deviation.

[0106] S3.6, monitor the values of the linkage performance indicator set and the port-industry linkage comprehensive index. When the value of the linkage comprehensive index exceeds the linkage threshold, a warning signal is triggered.

[0107] Further, the real-time values of each indicator in the linkage performance indicator set and the port-industry linkage comprehensive index result are continuously monitored. The port-industry linkage comprehensive index value is dynamically compared with the upper limit of the linkage threshold and the lower limit of the linkage threshold through real-time data stream processing technology. When the port-industry linkage comprehensive index value is continuously higher than the upper limit of the linkage threshold or continuously lower than the lower limit of the linkage threshold, an automatic warning signal is triggered.

[0108] S4, based on the performance indicators, use causal inference analysis technology to diagnose the bottleneck link and root cause of the linkage between port logistics and urban industry, and generate a diagnosis report.

[0109] S4.1, perform convergence cross mapping analysis on the port ship average time in port and enterprise raw material inventory days indicators in the linkage performance indicator set and the port-industry linkage comprehensive index, quantify the causal strength and direction between the port ship average time in port and enterprise raw material inventory days indicators, and calculate the causal contribution between the port ship average time in port and enterprise raw material inventory days using the convergence cross mapping algorithm.

[0110] The causal contribution expression is:

[0111] TE X→Y = H(Y future | Y past ) - H(Y future | Y past , X past );

[0112] wherein TE X→Y is the causal contribution degree, X is the average port ship time in port, Y is the enterprise raw material inventory days, Y future is the predicted enterprise raw material inventory days at the next time point, Y past is a set of predicted future enterprise raw material inventory days historical data, X past is the average port ship time in port historical data.

[0113] Further, the average port ship time in port sequence and the enterprise raw material inventory days time sequence in the linkage performance index set and the port-industry linkage comprehensive index are subjected to convergence cross mapping analysis, the shadow manifold of the average port ship time in port and the enterprise raw material inventory days is reconstructed, and the prediction accuracy based on the manifold is calculated, so as to quantify the causal influence strength and directionality of the average port ship time in port on the enterprise raw material inventory days, and the convergence cross mapping algorithm is used to calculate the causal contribution degree between the average port ship time in port and the enterprise raw material inventory days.

[0114] S4.2, based on the causal contribution degree, the average port ship time in port, the average speed of the port-dissipating road, and the industry raw material inventory turnover days index in the linkage performance index set that affect the port-industry linkage comprehensive index are identified, and the bottleneck link is formed.

[0115] Further, based on the causal contribution degree, the causal contribution degrees of all indexes in the linkage performance index set are sorted and screened, the average port ship time in port, the average speed of the port-dissipating road, and the industry raw material inventory turnover days index with the highest causal contribution degree are identified, and the average port ship time in port, the average speed of the port-dissipating road, and the industry raw material inventory turnover days index are formally determined as the bottleneck link affecting the port-industry linkage comprehensive index

[0116] S4.3, the bottleneck link is associated with the plan-actual comparison data set for correlation analysis, the specific operation event of the bottleneck link is located, and the root cause analysis result is obtained.

[0117] Further, the identified bottleneck link is associated with the plan-actual comparison data set for correlation analysis, the specific operation event causing the abnormal fluctuation of the bottleneck link is located through the time stamp matching and event tracing method, the decrease of the average speed of the port-dissipating road is associated with the traffic accident record or road maintenance operation record in a specific time period, and the detailed root cause analysis result is obtained.

[0118] S4.4, the bottleneck link, the causal strength, and the root cause analysis result are integrated to form the bottleneck link and root cause diagnosis report of the port logistics and city industry linkage.

[0119] Further, the identified bottleneck link list, causal strength values, and the specific job event root cause analysis results are integrated and compiled to generate a structured bottleneck link and root cause diagnosis report for port logistics and urban industry linkage.

[0120] S5. The supply chain demand instruction set is set as an optimization target, the resource data in the port logistics and urban industry supply chain data lake is set as a constraint, the bottleneck link and root cause are set as an optimization focus, and a multi-objective optimization model is constructed.

[0121] S5.1, based on the supply chain demand instruction set and the bottleneck link and root cause diagnosis report for port logistics and urban industry linkage, resource data is extracted from the port logistics and urban industry supply chain data lake.

[0122] Further, based on the product categories, demand quantity, and delivery time requirements specified in the supply chain demand instruction set, and in combination with the key bottleneck indicators identified in the bottleneck link and root cause diagnosis report for port logistics and urban industry linkage, corresponding resource data such as port berth quantity, shore crane efficiency, yard capacity, transportation fleet size, road capacity, production line capacity, and warehouse capacity is extracted from the resource data partition of the port logistics and urban industry supply chain data lake, providing complete input parameters for optimization model construction.

[0123] S5.2, by setting the supply chain demand instruction set as the target and the resource data as the constraint, the bottleneck mathematical tightening in the bottleneck link and root cause diagnosis report for port logistics and urban industry linkage is converted into a strengthened constraint, and a multi-objective optimization model is constructed.

[0124] Further, by setting the total demand satisfaction degree of the supply chain demand instruction set and the total cost minimization as the main optimization target, the extracted resource data is set as the capability constraint condition of the mathematical programming model, and the bottleneck link quantification characteristics in the bottleneck link and root cause diagnosis report for port logistics and urban industry linkage are converted into strengthened constraints through mathematical tightening mechanism, such as converting the bottleneck of insufficient road capacity in the diagnosis report into an artificial tightening constraint of the upper limit of the road capacity, thereby constructing a mixed integer programming model containing multi-objective functions, resource capability constraints, and bottleneck strengthened constraints.

[0125] S6, the multi-objective optimization model is solved to generate a supply chain optimization execution scheme.

[0126] S6.1, by performing non-dominated sorting between the population in the current evolution generation and the elite individual set generated in the last generation, an elite strategy is obtained.

[0127] Further, by merging the population individuals in the current evolution generation with the elite individual set generated in the last generation evolution process, the whole population after merging is sorted by the fast non-dominated sorting algorithm according to the objective function value, and the non-dominated solutions ranked in the front are screened out according to the crowding degree calculation, to form the elite strategy population of the new generation.

[0128] S6.2, the fast non-dominated sorting genetic algorithm with elite strategy is used to solve the multi-objective optimization model, to obtain the Pareto optimal solution set.

[0129] The expression of the Pareto optimal solution set is:

[0130]

[0131] Among them, is the Pareto optimal solution set, x is the decision variable vector, is the feasible solution set, x * is the Pareto optimal solution.

[0132] Further, the fast non-dominated sorting genetic algorithm with elite strategy is used to iteratively solve the multi-objective optimization model, the population is constantly updated through selection, crossover and mutation operations, and the diversity and convergence of the solution set are maintained by using non-dominated sorting and crowding degree comparison, and a Pareto optimal solution set close to the Pareto frontier is output, wherein each solution represents an optimization scheme that achieves the best trade-off between total cost, total time and bottleneck relief degree objectives.

[0133] S6.3, using the criteria decision analysis method, the final scheme is selected from the Pareto optimal solution set based on the cost, time efficiency and bottleneck relief degree preference weight, to form the supply chain optimization execution scheme.

[0134] Further, using the multi-criteria decision analysis method, based on the cost, time efficiency and bottleneck relief degree preference weight, the objective function values of all solutions in the Pareto optimal solution set are weighted, standardized and comprehensively scored, and the solution with the highest comprehensive score is selected as the final scheme, to form the supply chain optimization execution scheme.

[0135] S7, execute the supply chain optimization execution scheme, and feed back the execution effect data to the port logistics and city industry supply chain data lake, and update the multi-objective optimization model iteratively.

[0136] S7.1, the supply chain optimization execution scheme is issued to the port operation terminal, transportation scheduling terminal and enterprise production terminal to drive actual operation, and new port logistics data and city industry data generated by the actual operation of the supply chain optimization execution scheme are collected.

[0137] Further, the industry production instruction set and the port operation instruction set contained in the supply chain optimization execution scheme are respectively issued to the enterprise production execution terminal, the port operation terminal and the transportation management terminal through a standard data interface, to drive actual production, loading and unloading and transportation operations, and to collect new port logistics data and new city industry data such as actual port ship operation time, real-time inventory quantity of the yard, truck transportation track, actual production quantity of the production line and warehouse in-out record after the scheme execution through the Internet of Things sensing equipment and the business database interface.

[0138] S7.2, the port logistics new data and the city industry new data are transported back to the port logistics and city industry supply chain data lake for storage.

[0139] Further, the collected port logistics new data and city industry new data are transmitted to the real-time data partition of the port logistics and city industry supply chain data lake through a data pipeline, and are format-converted and partition-stored according to the original schema specification of the data lake, to complete the persistent feedback of the execution effect data.

[0140] S7.3, based on the historical data and the new data in the port logistics and city industry supply chain data lake, the parameters and weights of the multi-objective optimization model are recalibrated.

[0141] Further, based on the historical data accumulated in the port logistics and city industry supply chain data lake and the new port logistics data and city industry data, the Bayesian optimization method is used to recalibrate the target function weight coefficient and the constraint tightening parameter in the multi-objective optimization model, so that the multi-objective optimization model parameters are more suitable for the changes in the actual operation environment.

[0142] S7.4, a new generation of supply chain optimization execution scheme is generated by using the updated multi-objective optimization model, and a new round of optimization cycle is started.

[0143] Further, the multi-objective optimization model with recalibrated parameters is used to solve again, to generate a new generation of supply chain optimization execution scheme, and the scheme is put into practical application, so as to start a new round of closed loop cycle of data collection, feedback, parameter update and scheme optimization.

[0144] The embodiment also provides a computer device suitable for the case of the port logistics and city industry linkage supply chain optimization model construction method, which comprises a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the port logistics and city industry linkage supply chain optimization model construction method as described above.

[0145] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to communicate with external terminals in a wired or wireless manner. The wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0146] The embodiment also provides a storage medium having a computer program stored thereon, the program being executed by a processor to implement the method for constructing a supply chain optimization model for port logistics and city industry linkage as described above. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic storage, a flash memory, a magnetic disk or an optical disk.

[0147] To sum up, the present application constructs a supply chain data lake by collecting and fusing multi-source data, generates a demand instruction set based on consumption data prediction, calculates and monitors linkage performance indicators in real time, accurately diagnoses bottleneck links and root causes by using causal inference technology, constructs a multi-objective optimization model taking the instruction set as a target, resource data as a constraint and bottleneck as a focus, and solves the model to generate an execution scheme, realizes model iterative optimization through execution feedback, forms a closed-loop collaborative decision making driven by data, perceived by causality and continuously evolving, and effectively improves the accuracy, dynamics and overall efficiency of port and industry linkage.

[0148] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A method for constructing a supply chain optimization model that links port logistics with urban industries, characterized in that: It comprises collecting port logistics data, urban industry data and urban consumption data, cleaning and fusing processing to form a port logistics and urban industry supply chain data lake; Based on the urban consumption data in the port logistics and urban industry supply chain data lake, the consumption demand is predicted and inversely transformed into a supply chain demand instruction set for industrial production and port operation; According to the supply chain demand instruction set and the real-time data in the port logistics and urban industry supply chain data lake, the performance indicators of the port logistics and urban industry linkage are calculated and monitored, including the following steps: Extract the port logistics real-time data and urban industry real-time data from the port logistics and urban industry supply chain data lake, based on the historical operation cycle data, real-time ship AIS forecast data, enterprise production rhythm data and road traffic prediction data in the port logistics and urban industry supply chain data lake, the planning timeline in the supply chain demand instruction set is obtained through operational research; The planning timeline in the supply chain demand instruction set is accurately aligned with the time stamp of the port logistics real-time data and urban industry real-time data to form a plan-actual comparison data set; Based on the plan-actual comparison data set, the port end indicators, industry end indicators and linkage end indicators are extracted using time series pattern recognition and association rule mining technology, and the port end indicators, industry end indicators and linkage end indicators are summarized to form a linkage performance indicator set; The entropy weight method is used to assign weights to the indicators in the linkage performance indicator set, and the port-industry linkage comprehensive index is calculated; The port-industry linkage comprehensive index expression is: ； wherein, is the Hong Kong production linkage composite index, is the weight of the th indicator, is the standardized value of the th indicator in the current evaluation period, is the index of the indicator, is the total number of indicators. The linkage threshold is set by analyzing the historical linkage performance indicator set data in the port logistics and urban industry supply chain data lake; The values of the linkage performance indicator set and the port-industry linkage comprehensive index are monitored, and when the value of the linkage comprehensive index exceeds the linkage threshold, a warning signal is triggered; Based on the performance indicators, the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage are diagnosed using causal inference analysis technology, including the following steps: Convergent cross mapping analysis is performed on the port ship average time in port and enterprise raw material inventory days indicators in the linkage performance indicator set and the port-industry linkage comprehensive index, the causal strength and direction between the port ship average time in port and enterprise raw material inventory days indicators are quantified, and the causal contribution degree between the port ship average time in port and enterprise raw material inventory days is calculated using convergent cross mapping algorithm; Based on the causal contribution degree, the port ship average time in port, the average speed of the port access road and the industry raw material inventory turnover days indicators in the linkage performance indicator set which affect the port-industry linkage comprehensive index are identified, forming the bottleneck link; The bottleneck link and the plan-actual comparison data set are associated to locate the specific operation event of the bottleneck link, and the root cause analysis result is obtained; The bottleneck link, causal strength and root cause analysis result are integrated to form the bottleneck link and root cause diagnosis report of the port logistics and urban industry linkage. The supply chain demand instruction set is taken as an optimization target, resource data in the port logistics and urban industry supply chain data lake is taken as a constraint, a bottleneck link and a root are taken as an optimization focus, a multi-objective optimization model is constructed, the multi-objective optimization model is solved, and a supply chain optimization execution scheme is generated; The supply chain optimization execution scheme is executed, and execution effect data is fed back to the port logistics and urban industry supply chain data lake, and the multi-objective optimization model is iteratively updated. 2.The method of claim 1, wherein the method is characterized in that: Port logistics data, urban industry data and urban consumption data are collected, cleaned and fused to form a port logistics and urban industry supply chain data lake, including the following steps, Port logistics data, urban industry data and urban consumption data are obtained by acquiring port logistics data, urban industry data and urban consumption data through a sensing device and a data interface, and performing deduplication, error correction and abnormal value processing on the port logistics data, urban industry data and urban consumption data; The cleaned port logistics data, urban industry data and urban consumption data are converted into a unified space-time format and measurement unit to obtain standardized port logistics data, urban industry data and urban consumption data; The standardized port logistics data, urban industry data and urban consumption data are aligned and associated according to entities and time to obtain fused port logistics data, urban industry data and urban consumption data; The fused port logistics data, urban industry data and urban consumption data are persistently stored to form a port logistics and urban industry supply chain data lake.

3. The method of claim 2, wherein the method is characterized by: Based on the urban consumption data in the port logistics and urban industry supply chain data lake, consumption demand prediction is performed and is inversely transformed into a supply chain demand instruction set for industry production and port operation, including the following steps, Urban consumption data in a one-year time range is extracted from the port logistics and urban industry supply chain data lake; A city consumption demand prediction report is generated by using a long short-term memory network model to train derived features constructed based on the city consumption data using an association rule mining method; The maximum operation rate of port equipment, the saturation capacity of the yard, the upper limit of road traffic capacity, the maximum load of the production line and the size of the transportation fleet data are collected, cleaned and standardized to form the capacity and transportation capacity constraints in the port logistics and urban industry supply chain data lake; The city consumption demand prediction report is taken as an optimization target, combined with the capacity and transportation capacity constraints in the port logistics and urban industry supply chain data lake, and a mathematical programming is solved to obtain an industry production instruction set and a port operation instruction set; The industry production instruction set and the port operation instruction set are combined to form a supply chain demand instruction set.

4. The method of claim 3, wherein the method is characterized in that: The supply chain demand instruction set is taken as an optimization target, resource data in the port logistics and urban industry supply chain data lake is taken as a constraint, a bottleneck link and a root are taken as an optimization focus, a multi-objective optimization model is constructed, including the following steps, Based on the supply chain demand instruction set and the bottleneck link and root diagnosis report of the port logistics and urban industry linkage, resource data is extracted from the port logistics and urban industry supply chain data lake; By targeting the supply chain demand instruction set, resource data as constraints, the bottleneck link and root cause diagnosis report of port logistics and urban industry linkage are transformed into strong constraints by bottleneck mathematics tightening, and a multi-objective optimization model is constructed.

5. The method of claim 4, wherein the method is characterized in that: Solving the multi-objective optimization model generates a supply chain optimization execution plan, including the following steps, Through non-dominated sorting between the current evolutionary generation population and the elite individual set generated by the last generation, an elite strategy is obtained; The fast non-dominated sorting genetic algorithm with elite strategy is used to solve the multi-objective optimization model to obtain a Pareto optimal solution set; Using the criteria decision analysis method, the final scheme is selected from the Pareto optimal solution set based on the cost, time efficiency, and bottleneck relief degree preference weight, forming the supply chain optimization execution plan.

6. The method of claim 5, wherein, Execute the supply chain optimization execution plan and feed back the execution effect data to the port logistics and urban industry supply chain data lake, and update the multi-objective optimization model, including the following steps, The supply chain optimization execution plan is issued to the port operation terminal, transportation scheduling terminal and enterprise production terminal to drive actual operation, and new port logistics data and urban industry data generated by the actual operation of the supply chain optimization execution plan are collected; The new port logistics data and urban industry data are transported back to the port logistics and urban industry supply chain data lake for storage; Based on the historical data and new data in the port logistics and urban industry supply chain data lake, the parameters and weights of the multi-objective optimization model are recalibrated; Using the updated multi-objective optimization model to generate a new generation of supply chain optimization execution plan starts a new round of optimization cycle.

7. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that: The processor executes the computer program to realize the steps of the supply chain optimization model construction method of claim 1-6.

8. A computer readable storage medium having stored thereon a computer program, characterized in that: The computer program is executed by the processor to realize the steps of the supply chain optimization model construction method of claim 1-6.

Images