Port throughput analog simulation model

Abstract

The invention provides a port throughput analog simulation model. According to the method, a technical system including dynamic parameter self-calibration, multi-link dynamic simulation and uncertainty quantification is innovatively constructed, historical throughput, ship AIS and multimodal combined transport data are fused, and through a reinforcement learning closed-loop mechanism (automatic optimization with deviation exceeding 5%) and a Monte Carlo algorithm, full-process workload coupling calculation and wharf-division type throughput generation are realized, so that the throughput of the ship can be quickly and accurately calculated. The prediction precision (MAPE < = 2.1%) is improved, the operation efficiency is optimized (the ship in-port time is shortened by 37.5%, and the storage yard turnover rate is improved by 22%), the green development (the carbon emission intensity is reduced by 15%) and the emergency decision (the fluctuation is predicted 72 hours in advance) are supported, and the reduction of the port investment is assisted.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of port throughput simulation, in particular to a port throughput simulation model. BACKGROUND

[0002] The continuous growth of global trade volume drives the evolution of ports into a composite hub. By 2025, the proportion of 20,000 TEU class super large container ships operating globally has reached 35%, which puts higher requirements on the accuracy of port throughput prediction and the optimization capability of port operation. The existing models have the following problems: (1) Insufficient dynamic adaptation: existing models are mostly based on fixed parameter static simulation, which cannot respond to dynamic variables in port operation in real time, such as changes in the frequency of arrival of super large ships, sudden equipment failures, etc., resulting in a deviation rate of more than 25% between simulation results and actual operation; (2) Lack of multi-system coordination: traditional models mostly focus on a single operation link, ignoring the dynamic interaction of ports with railway, highway and other intermodal systems. The efficiency loss caused by the transfer distance of more than 50 kilometers in rail-sea intermodal transportation is not described, and the model cannot support comprehensive logistics decision optimization; (3) Insufficient quantification of uncertainty: existing simulations mostly avoid random interference factors, such as FlexSim simulation assuming no equipment failure and no damage to goods, and lack the ability to quantitatively evaluate sudden factors such as extreme weather and policy adjustments, resulting in a simulation credibility of less than 60% in emergency scenarios; (4) Lack of integration of green indicators: there is a lack of quantitative modules for green parameters such as the berthing efficiency of new energy ships and the load of shore power facilities, which cannot meet the needs of port sustainable development planning; (5) Parameter iteration lag: traditional models rely on offline calibration of historical data. When the annual growth rate of port throughput reaches 7.2%, the parameter update lag results in a prediction error of more than 15%, making it difficult to support dynamic resource allocation decisions. SUMMARY

[0003] The present application aims to at least solve one of the technical problems existing in the prior art, and provides a port throughput simulation model.

[0004] To achieve the above-mentioned purpose, the present application adopts the following technical solutions: In a first aspect, the present application provides a port throughput simulation model, comprising: A database building module is used to integrate historical throughput data, ship AIS dynamic data, equipment operation log data, intermodal scheduling data and meteorological statistical data to form a basic database; A data preprocessing module is used to detect outliers in the basic database using an improved particle swarm algorithm, fill in missing values by linear interpolation, and standardize different dimension data; A parameter matrix construction module is configured to form a three-dimensional parameter matrix including basic parameters, system interference parameters, manual compensation parameters, and operation influence parameter matrix and random event library parameters by disassembling key influence factors of a port operation whole process; A dynamic parameter self-calibration module is configured to form a "deviation feedback-parameter correction-simulation iteration" closed loop mechanism based on reinforcement learning, collect ship AIS data and equipment IoT state in real time, dynamically update simulation parameters, and automatically trigger parameter optimization when the actual throughput and the simulation throughput deviation is more than 5%; A multi-link dynamic simulation module is configured to calculate the operation amount of each link of anchorage turnaround, wharf front, horizontal transportation, yard operation and lock passage based on the three-dimensional parameter matrix coupling, generate link operation amount data of each wharf type, weight and sum the link operation amount data of each wharf type based on the operation influence parameter matrix to generate wharf type throughput, and aggregate the wharf type throughput to generate a total port throughput; An interference and compensation coupling module is configured to quantify the influence of TOS system and ECS system interruption on the efficiency of each link of port operation, and generate compensated operation amount data in combination with manual intervention coefficient and manual replacement efficiency; An uncertainty quantification simulation module is configured to construct a random event library including multiple types of interference factors, output throughput prediction results of different confidence intervals based on a probability model trained by historical data and a Monte Carlo algorithm, and realize uncertainty quantification.

[0005] In some possible embodiments, a dynamic berthing planning module is further included, which is connected with the parameter matrix construction module and the multi-link dynamic simulation module. The dynamic berthing planning module is configured to receive the wharf front operation amount data output by the multi-link dynamic simulation module, and optimize a 150-dimensional coded ship berthing scheme based on a genetic algorithm while taking berth capacity and crane quantity as constraint conditions.

[0006] In some possible embodiments, a multimodal transport scheduling module is further included, which is connected with the database building module and the interference and compensation coupling module. The multimodal transport scheduling module is configured to extract the multimodal transport scheduling data from the database building module and integrate an information sharing platform to realize real-time synchronization of port, railway and highway data, plan an optimal transfer path based on an ant colony algorithm, and output path optimization results.

[0007] In some possible embodiments, a green simulation module is further included, which is connected with the parameter matrix construction module and the uncertainty quantification simulation module, and is configured to construct a regression model of shore power utilization rate and throughput by combining the shore power coverage rate parameter extracted from the parameter matrix construction module and the anchorage turnaround operation data, the wharf front operation data, the horizontal transportation operation data, the yard operation data and the lock passage volume data obtained from the multi-link dynamic simulation module; and to quantitatively predict the change range of throughput and the amount of carbon emission reduction when the shore power coverage rate is increased from 20% to 50% based on the regression model of shore power utilization rate and throughput.

[0008] In some possible embodiments, the operation influence coefficient matrix includes an anchorage turnaround influence coefficient on throughput a wharf operation influence coefficient on throughput a horizontal transportation influence coefficient on throughput a yard operation influence coefficient on throughput a lock passage influence coefficient on throughput and satisfies .

[0009] In some possible embodiments, the sub-wharf type link operation data includes anchorage turnaround operation data, wharf front operation data, horizontal transportation operation data, yard operation data and lock passage volume data; wherein: the calculation formula of the anchorage turnaround operation data is: , wherein, is the anchorage capacity (unit: ship), is the average waiting time of a ship (unit: hour), is the interference coefficient of TOS on anchorage (value: 0-1), is the availability of TOS system (unit: %), is the average handling volume of a ship (unit: TEU / ship), is the pilot efficiency coefficient, is the anchorage scheduling operation coefficient, is the manual intervention coefficient, is the manual replacement TOS efficiency (x=1, 2, 3, x=1, 0.7-0.9; x=2, 0.5-0.7; x=3, 0.2-0.4), is the manual intervention response time (unit: hour), is the total duration of the statistical period (unit: hour); the calculation formula of the wharf front operation data is: , , is the berth operation time (unit: hour), is the berth operation time (unit: hour), is the number of quayside cranes (unit: table), is the quayside crane operation efficiency (unit: TEU / hour*table), is the average operation time of quayside cranes (unit: hour), is the terminal operation coefficient, is the ECS availability (unit: %), is the TOS interference attenuation coefficient (value 0-1), is the TOS dependence (x=1, 2, 3, x=1 0.3-0.5, x=2 0.6-0.8, x=3 0.8-1.0), is the ECS interference attenuation coefficient (value 0-1, x=1 0), is the ECS dependence (x=1, 2, 3, x=1 0, x=2 0.5-0.7, x=3 0.9-1.0), is the artificial replacement ECS efficiency (x=1, 2, 3, x=1 0, x=2 0.4-0.6, x=3 0.1-0.3); The calculation formula of horizontal transportation operation volume data is: , , is the number of horizontal transportation equipment (unit: table), is the average operation efficiency of the equipment (unit: TEU / hour*table), is the average daily operation time of the equipment (unit: hour), is the transportation path smoothness coefficient, is the ECS interference path attenuation coefficient (value 0-1), is the ECS control dependence (x=1, 2, 3, x=1 0, x=2 0.4-0.6, x=3 0.7-1.0), is the horizontal transportation operation coefficient; The calculation formula of yard operation volume data is: , , is the amount of stored goods in the yard (unit: TEU), is the total amount of stored goods in the yard (unit: TEU), is the number of yard cranes (unit: table), is the yard crane operation efficiency (unit: TEU / hour*table), is the average operation time of the yard bridge (unit: hour), is the operation coefficient of the yard, is the interference coefficient of the TOS to the yard (value 0-1), is the TOS planning dependence (x = 1, 2, 3, x = 1 0.2-0.4, x = 2 0.5-0.7, x = 3 0.8-1.0), is the interference coefficient of the ECS to the yard (value 0-1, x = 1 0), is the ECS yard bridge dependence (x = 1, 2, 3, x = 1 0, x = 2 0.3-0.5, x = 3 0.8-1.0); The calculation formula of the gate traffic volume data is: , In the formula, is the gate traffic efficiency (unit: vehicle / hour), is the number of gate channels (unit: pieces), is the gate traffic time (unit: hour), is the average capacity of the container truck (unit: TEU / vehicle), is the operation coefficient of the gate, is the interference coefficient of the TOS to the gate (value 0-1), is the dependence of the TOS to the gate (x = 1, 2, 3, x = 1 0.3-0.5, x = 2 0.6-0.8, x = 3 0.8-1.0).

[0010] In some possible embodiments, the calculation formula of the sub-port type throughput is: , In the formula, is the total throughput of the xth type of port (including interference and compensation), x = 1, 2, 3 respectively corresponds to a traditional port, a semi-automated port, and a fully automated port, is the anchorage turnaround operation volume data of the xth type of port, is the pre-terminal operation volume data of the xth type of port, is the horizontal transportation operation volume data of the xth type of port, is the yard operation volume data of the xth type of port, is the gate traffic volume data of the xth type of port; The calculation formula of the total port throughput is: , in which, is the throughput of a traditional port, is the throughput of a semi-automated port, is the throughput of a fully automated port.

[0011] In some possible embodiments, a model verification and calibration module is further included, which is connected with the multi-link dynamic simulation module and the dynamic parameter self-calibration module respectively, and is used for verifying and optimizing model precision, comprising: Historical back test: historical port actual operation data is used to verify the output result of the multi-link dynamic simulation module, so as to ensure that the RMSE value is less than 8.6% and the MAPE value is less than or equal to 2.1%; Sensitivity analysis: key parameters are identified, a parameter adjustment threshold system is established, and the dynamic parameter self-calibration module is triggered when the threshold is exceeded; Field test: 120 ship measurement data are collected in a coastal port for 3 months of trial operation, and the model parameters are calibrated.

[0012] In a second aspect, the present application provides an electronic device, comprising: One or more processors; A storage unit for storing one or more programs, which can enable the one or more processors to implement the port throughput simulation model as described above when the one or more programs are executed by the one or more processors.

[0013] In a third aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program can implement the port throughput simulation model as described above when executed by a processor.

[0014] The port throughput simulation model of the embodiments of the present application has the following beneficial effects: 1. The prediction accuracy of the present application is significantly improved. The dynamic optimization of simulation accuracy is realized through the synergistic effect of the dynamic parameter self-calibration module and the model verification and calibration module, a closed-loop mechanism is constructed based on reinforcement learning (deviation is automatically optimized when the deviation is greater than 5%), the parameter response delay is less than or equal to 10 minutes, the deviation is stably controlled within 5%, the average absolute percentage error (MAPE) of prediction is less than or equal to 2.1%, which is reduced by 63.8% compared with the grey prediction model, and the deviation rate is less than or equal to 3.5% when the proportion of ultra-large ships increases by 10% (the traditional model is 18.7%).

[0015] 2. The operation efficiency optimization effect of the present application is quantified. The present application relies on multi-link dynamic simulation module, dynamic berthing planning module and multimodal transport scheduling module to realize full-process efficiency breakthrough; the multimodal transport scheduling module integrates information sharing platform, optimizes the path through ant colony algorithm, shortens the average time of ships in port from 72 hours to 45 hours, reduces by 37.5%, far exceeds the optimization effect of 15% of the dynamic port capacity model, and reduces the return empty running rate from 41% to 22%; the multi-link dynamic simulation module couples the calculation of yard operation quantity to promote the yard turnover rate by 22%, when the yard utilization rate reaches 82%, the turnover time growth elasticity coefficient is reduced from 1.34 to 0.89, effectively alleviating the congestion amplification effect; the dynamic berthing planning module optimizes the 150-dimensional coding berthing scheme based on genetic algorithm, and the average waiting time of ships is shortened by 25%.

[0016] 3. The green development support ability of the present application is outstanding. The present application constructs the regression model of shore power utilization rate and throughput through the green simulation module, which can accurately quantify the comprehensive benefits of green measures; through optimizing the shore power use and new energy ship scheduling scheme, the port can increase the throughput by 12.3% while reducing the carbon emission intensity by 15%, realizing the coordinated optimization of efficiency and environmental protection, and solving the problem that the traditional model cannot quantify the green benefits, resulting in the evaluation deviation of the green investment return rate of some ports being more than 40%, while the present application can accurately predict the throughput variation range and carbon emission reduction amount during the process of increasing the shore power coverage rate from 20% to 50%.

[0017] 4. The emergency response ability of the present application is comprehensively enhanced. The present application improves the extreme scenario response ability based on the uncertainty quantification simulation module and the interference and compensation coupling module; through the uncertainty quantification simulation module, a random event library containing multiple interference factors is established, and based on the Monte Carlo algorithm, the throughput prediction results of different confidence intervals are output, the simulation credibility of emergency scenarios is improved from less than 60% of the traditional model to more than 85%, and when the port is temporarily closed due to extreme weather, the throughput fluctuation range can be predicted 72 hours in advance, which is helpful for assisting in formulating emergency scheduling scheme; the interference and compensation coupling module quantifies the influence of TOS / ECS system interruption, combines with the artificial intervention coefficient and the artificial replacement efficiency (the ECS efficiency of fully automated wharf is 0.1-0.3), and through the dynamic integration of system interruption quantification results and artificial intervention parameters, the precise response and operation process adjustment to system abnormal scenarios are realized.

[0018] 5. The decision support value of the present application is significant. The present application supports multi-dimensional decision-making through the parameter matrix construction module and the multi-link dynamic simulation module. For example, in the port expansion planning, by simulating different berth quantity schemes, the optimal expansion scale is determined to be 3 100,000-ton berths, which reduces the investment by 120 million yuan compared with the initial scheme, and meets the long-term demand of 40% throughput growth. Through subsequent operation verification, the resource utilization rate is improved by 28%, the equipment idle rate is reduced by 15%, and the investment recovery period is shortened by 1.5 years; at the same time, it supports the throughput simulation of different types of wharfs (traditional / semi-automatic / fully automatic), and provides data support for intelligent transformation. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The structural schematic diagram of the example electronic device for realizing the port throughput simulation model of the present application is shown in FIG. 1. Figure 2 The structural schematic diagram of the port throughput simulation model of the present application is shown in FIG. 2. DETAILED DESCRIPTION

[0020] The technical solutions of the present application will be described clearly and completely below with reference to the drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0021] Figure 1 The structural schematic diagram of the example electronic device for realizing the port throughput simulation model of the present application is shown in FIG. 1. Figure 1 As shown in FIG. 1, the electronic device 100 includes one or more processors 110, one or more storage devices 120, one or more input devices 130, one or more output devices 140, etc., and these components are interconnected through a bus system 150 and / or other forms of connection mechanism. It should be noted that, Figure 1 The components and structure of the electronic device shown are only exemplary and are not limiting, and the electronic device can also have other components and structures as needed.

[0022] The processor 110 can be a central processing unit (CPU), or can be other forms of processing units composed of multiple processing cores, or having data processing capability and / or instruction execution capability, and can control other components in the electronic device 100 to perform desired functions.

[0023] The storage 120 can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), cache memory, and / or the like. The non-volatile memory, for example, can include read only memory (ROM), hard disk, flash memory, and / or the like. One or more computer program instructions can be stored on the computer-readable storage media, which the processor can execute to implement the client functions (implemented by the processor) in the embodiments of the present disclosure described below and / or other desired functions. Various application programs and various data, such as various data used and / or generated by the application programs, and the like, can also be stored in the computer-readable storage media.

[0024] The input device 130 can be a device used by a user to input instructions, and can include one or more of a keyboard, a mouse, a microphone, a touch screen, and the like.

[0025] The output device 140 can output various information (such as an image or a sound) to the outside (such as a user), and can include one or more of a display, a speaker, and the like.

[0026] Figure 2 The structure diagram of the port throughput simulation model of the present application. As shown in Figure 2As shown, a port throughput simulation model includes a database building module 201: for integrating historical throughput data (for example, historical throughput data from 2015 to 2024), ship AIS dynamic data, equipment operation log data, multimodal transport scheduling data, and meteorological statistical data to form a basic database (10 TB level); a data preprocessing module 202: for detecting abnormal values of the basic database using an improved particle swarm algorithm, filling in missing values through linear interpolation, and standardizing different dimension data; a parameter matrix construction module 203: for forming a three-dimensional parameter matrix containing basic parameters, system disturbance parameters, manual compensation parameters, operation influence parameter matrix, and random event library parameters by disassembling key influencing factors of the whole process of port operation; a dynamic parameter self-calibration module 204: for constructing a "deviation feedback-parameter correction-simulation iteration" closed loop mechanism based on reinforcement learning, dynamically updating simulation parameters by real-time collection of ship AIS data and equipment IoT state (sampling frequency ≥ 1 time / minute), automatically triggering parameter optimization when the deviation between actual throughput and simulation throughput is more than 5%, and parameter response delay ≤ 10 minutes; a multi-link dynamic simulation module 205: for calculating the operation amount of each link of anchorage turnaround, wharf front, horizontal transportation, yard operation, and lock passage based on the three-dimensional parameter matrix coupling, generating link operation amount data of each wharf type, weighting and summing the link operation amount data of each wharf type based on the operation influence parameter matrix to generate the throughput of each wharf type, and aggregating the throughput of each wharf type to generate the total port throughput; an interference and compensation coupling module 206: for quantifying the impact of TOS system (i.e. wharf production operation system) and ECS system (i.e. equipment control system) interruption on the efficiency of each link of port operation, and generating compensated operation data combined with manual intervention coefficient and manual replacement efficiency; an uncertainty quantification simulation module 207: for constructing a random event library containing multiple types of interference factors, outputting throughput prediction results with different confidence intervals based on the probability model trained by historical data and the Monte Carlo algorithm, realizing uncertainty quantification, and improving the credibility of emergency scenario simulation to more than 85%.

[0027] Specifically, the key influencing factors of the whole process of port operation include but are not limited to anchorage scheduling, wharf front, horizontal transportation, yard operation, and lock passage.

[0028] Specifically, the improved particle swarm algorithm is used to detect abnormal values of the basic database, which specifically includes the following steps: initializing the particle swarm, defining the fitness function, updating the particle position and velocity, abnormal value detection and processing, and termination condition.

[0029] Specifically, the "deviation feedback-parameter correction-simulation iteration" closed loop mechanism is constructed based on reinforcement learning, which specifically includes the following steps: defining the state space, defining the action space, defining the reward function, Q-learning update, and simulation iteration.

[0030] Specifically, the multiple interference factors include, but are not limited to, natural factors such as extreme weather, hardware and software device failure factors, human factors such as strikes and operation errors, technical factors such as network attacks and data loss, and policy and market factors.

[0031] In some embodiments, a dynamic berthing planning module is further included, which is connected with the parameter matrix construction module and the multi-link dynamic simulation module. The dynamic berthing planning module is used to receive the wharf front operation data output by the multi-link dynamic simulation module, and optimize the 150-dimensional coded ship berthing scheme based on a genetic algorithm while taking the berth capacity and the number of cranes as constraint conditions, so as to calculate the optimal berthing sequence and shorten the average waiting time of the ship by 25%.

[0032] Specifically, the 150-dimensional coded ship berthing scheme is optimized based on a genetic algorithm while taking the berth capacity and the number of cranes as constraint conditions, which specifically includes the following steps: coding and initialization, fitness function, selection, crossover and mutation, constraint processing, termination and output.

[0033] In some embodiments, a multimodal transport scheduling module is further included, which is connected with the database building module and the interference and compensation coupling module. The multimodal transport scheduling module is used to extract multimodal transport scheduling data from the database building module and integrate an information sharing platform to realize real-time synchronization of port, railway and highway data, plan an optimal transfer path based on an ant colony algorithm, and output the path optimization result, which can reduce the empty running rate of the return trip from 41% to 22%.

[0034] Specifically, the information sharing platform is the core data exchange infrastructure of the multimodal transport scheduling module, which aims to solve the "data island" problem between different subjects such as ports and railways, and ensure real-time synchronization and sharing of logistics data.

[0035] Specifically, the ant colony algorithm is used to plan the optimal transfer path, which specifically includes the following steps: problem modeling, algorithm initialization, path construction, pheromone update, termination and output.

[0036] In some embodiments, a green simulation module is further included, which is connected with the parameter matrix construction module and the uncertainty quantification simulation module. The green simulation module is used to combine the shore power coverage rate parameters extracted from the parameter matrix construction module and the anchorage turnaround operation data, wharf front operation data, horizontal transportation operation data, yard operation data and lock passage volume data obtained from the multi-link dynamic simulation module, construct a regression model of shore power utilization rate and throughput, and quantitatively predict the throughput variation amplitude and carbon emission reduction amount based on the regression model of shore power utilization rate and throughput when the shore power coverage rate is increased from 20% to 50%.

[0037] Specifically, a regression model of shore power utilization rate and throughput is constructed, specifically including the following steps: variable definition and data preparation, regression model establishment, model training and verification, prediction and application.

[0038] In some embodiments, the job influence coefficient matrix includes an anchorage turnaround influence coefficient on throughput , a wharf operation influence coefficient on throughput , a horizontal transportation influence coefficient on throughput , a yard operation influence coefficient on throughput , a lock passage influence coefficient on throughput , and satisfies .

[0039] In some embodiments, the per-wharf-type link operation amount data includes anchorage turnaround operation amount data, wharf front operation amount data, horizontal transportation operation amount data, yard operation amount data, and lock passage amount data; wherein: The calculation formula of the anchorage turnaround operation amount data is: , In the formula, is the anchorage capacity (the maximum number of ships that the anchorage can accommodate at the same time, unit: ship), is the average waiting time of the ship in the anchorage (the average waiting time of the ship in the anchorage, unit: hour), is the interference coefficient of TOS to the anchorage (the decay ratio of TOS interference to the scheduling efficiency of the anchorage, the value is 0-1), is the availability of TOS system (the proportion of normal working time of the terminal operation system, unit: %), is the average handling capacity of the ship (the average handling capacity of a single ship, unit: TEU / ship), is the pilot efficiency coefficient (the influence coefficient of pilot operation on ship turnaround efficiency), is the anchorage scheduling operation coefficient (the overall efficiency reduction coefficient of anchorage operation), is the manual intervention coefficient (the efficiency coefficient of manual intervention when the system is interrupted), is the efficiency of manual TOS replacement (the efficiency of manual TOS replacement, x=1, 2, 3, x=1 0.7-0.9; x=2 0.5-0.7; x=3 0.2-0.4), is the manual intervention response time (the delay time from the system terminal to the manual intervention, unit: hour), is the total duration of the statistical period (the time period of throughput statistics, unit: hour) ; in the non-interference scenario, =1 (TOS is completely normal), and = 0 (no manual intervention); in the interference scenario, TOS failure (x = 1, 2, 3, x = 1 when 0.3-0.5, x = 2 when 0.6-0.8, x = 3 when 0.8-1.0) Under the interference scenario, TOS failure (x = 1, 2, 3, x = 1 when 0.3-0.5, x = 2 when 0.6-0.8, x = 3 when 0.8-1.0) will prolong the waiting time (the denominator increases), resulting in a decrease in the work load; after manual intervention, the work load is partially restored by shortening the waiting time (the numerator compensates). The calculation formula of the wharf front work load data is: , In the formula, is the berth operation time (the actual time of the berth for loading and unloading (excluding idle / maintenance time), unit: hour), is the berth operation time (the total available time of the berth in the statistical period (including operation and non-operation time), unit: hour), is the number of shore cranes (the number of shore cranes of each type of wharf, unit: units), is the shore crane operation efficiency (the number of containers handled per hour by a single shore crane, unit: TEU / hour*unit), is the average operation time of the shore crane (the average operation time of a single shore crane per day, unit: hour), is the wharf operation coefficient (1: traditional, 2: semi-automatic, 3: fully automatic), is the ECS availability (the proportion of normal operation time of the equipment control system, unit: %), is the TOS interference attenuation coefficient (the attenuation ratio of the wharf front operation efficiency when TOS is interrupted, value 0-1), is the TOS dependency (the dependency of the wharf front operation on TOS, x = 1, 2, 3, x = 1 when 0.3-0.5, x = 2 when 0.6-0.8, x = 3 when 0.8-1.0), is the ECS interference attenuation coefficient (the attenuation ratio of the wharf front operation efficiency when ECS is interrupted, value 0-1, x = 1 when 0), is the ECS dependency (the dependency of the wharf front operation on ECS, x = 1, 2, 3, x = 1 when 0, x = 2 when 0.5-0.7, x = 3 when 0.9-1.0), is the manual replacement ECS efficiency (the efficiency of manual replacement of ECS functions, x = 1, 2, 3, x = 1 when 0, x = 2 when 0.4-0.6, x = 3 when 0.1-0.3); The calculation formula of the horizontal transportation work load data is: , In the formula, is the number of horizontal transportation devices (the total number of horizontal transportation devices, unit: units), is the average operation efficiency of the device (the number of containers transferred per hour by a single device, unit: TEU / hour*unit), is the daily average operation time of the equipment (the effective operation time of each device per day, unit: hour), is the transport path unobstructed coefficient (the reduction coefficient of transport efficiency due to path congestion), is the ECS interference path attenuation coefficient (the additional attenuation ratio of the path unobstructed coefficient when the ECS is interrupted, the value is 0-1), is the ECS control dependency (the dependency of horizontal transportation on ECS, x = 1, 2, 3, x = 1 is 0, x = 2 is 0.4-0.6, x = 3 is 0.7-1.0), is the horizontal transportation operation coefficient (horizontal transportation overall efficiency reduction coefficient); The calculation formula of the yard operation amount data is: , In the formula, is the yard storage cargo amount (the actual number of containers stored in the current yard, unit: TEU), is the total yard storage cargo amount (the maximum design storage capacity of the yard, unit: TEU), is the number of yard bridges (the number of yard bridges actually participating in operation, unit: sets), is the yard bridge operation efficiency (unit: TEU / hour*set), is the average operation time of the yard bridge (the average operation time of each yard bridge per day, unit: hour), is the yard operation running coefficient (the overall efficiency reduction coefficient of the yard operation), is the TOS interference coefficient to the yard (the attenuation ratio of the yard efficiency when the TOS is interrupted, the value is 0-1), is the TOS planning dependency (the dependency of the yard on the TOS plan, x = 1, 2, 3, x = 1 is 0.2-0.4, x = 2 is 0.5-0.7, x = 3 is 0.8-1.0), is the ECS interference coefficient to the yard (the attenuation ratio of the yard efficiency when the ECS is interrupted, the value is 0-1, x = 1 is 0), is the ECS yard bridge dependency (the dependency of the yard bridge on the ECS, x = 1, 2, 3, x = 1 is 0, x = 2 is 0.3-0.5, x = 3 is 0.8-1.0); The calculation formula of the gate traffic volume data is: , In the formula, is the gate traffic efficiency (the number of vehicles released per hour for a single channel, unit: vehicles / hour), is the number of gate channels (the number of channels for the port traffic container trucks, unit: pieces), is the gate traffic time (the total opening time of the gate in the statistical period, unit: hour), is the average capacity of container trucks (the average number of standard containers carried by a single truck, unit: TEU / vehicle), is the gate operation coefficient (overall efficiency reduction coefficient of gate operation), is the interference coefficient of TOS to the gate (the attenuation ratio of the efficiency of the gate when TOS is interrupted, the value is 0-1), is the dependence of TOS on the gate (the dependence of the gate on TOS, x=1, 2, 3, x=1 0.3-0.5, x=2 0.6-0.8, x=3 0.8-1.0).

[0040] In some embodiments, the calculation formula of the throughput of the wharf type is: , In the formula, is the total throughput of the xth wharf (including interference and compensation), x=1, 2, 3 respectively corresponding to a traditional wharf, a semi-automated wharf, and a fully automated wharf, is the anchorage turnover operation data of the xth wharf (total anchorage ship transfer volume (including TOS interference and manual compensation)), is the wharf front operation data of the xth wharf (total wharf front loading and unloading volume (including system interference and manual compensation)), is the horizontal transportation operation data of the xth wharf (total horizontal transportation transfer volume (including ECS interference and manual compensation)), is the yard operation data of the xth wharf (total yard operation volume (including system interference and manual compensation)), is the gate passage data of the xth wharf (total gate passage volume (including TOS interference and manual compensation)); The calculation formula of the total throughput of the port is: In the formula, is the throughput of the traditional wharf, is the throughput of the semi-automated wharf, is the throughput of the fully automated wharf.

[0041] In some embodiments, a model verification and calibration module is further included, which is connected with the multi-link dynamic simulation module and the dynamic parameter self-calibration module respectively, for verifying and optimizing the model accuracy, including: Historical backtest: using historical port actual operation data (for example, data from 2020 to 2023), verifying the output results of the multi-link dynamic simulation module, ensuring that the RMSE value is <8.6% and the MAPE value is ≤2.1%; Sensitivity analysis: identifying key parameters, establishing a parameter adjustment threshold system, and triggering the dynamic parameter self-calibration module when the threshold is exceeded; Field test: In a coastal port, 3 months of trial operation, 120 ship data collection, model parameter calibration.

[0042] Specifically, the key parameters include but are not limited to operation influence coefficient (anchorage turnover, wharf front, horizontal transportation, yard operation, lock passage), system dependence parameter (TOS, ECS), core efficiency parameter (shore crane operation efficiency, yard crane operation efficiency, horizontal transportation equipment average operation efficiency), interference and compensation parameter (manual TOS efficiency substitution, manual ECS efficiency substitution) and operation coefficient (anchorage scheduling operation coefficient, wharf operation coefficient).

[0043] Specifically, the parameter adjustment threshold system is established, which specifically includes the following steps: sensitivity quantization, threshold level setting, dynamic monitoring and triggering, and closed-loop feedback.

[0044] The port throughput simulation model of the embodiment of the present application has the following advantages: 1. The prediction accuracy of the present application is significantly improved. The present application realizes dynamic optimization of simulation accuracy through the synergistic effect of the dynamic parameter self-calibration module and the model verification and calibration module, establishes a closed-loop mechanism based on reinforcement learning (deviation over 5% automatic optimization), makes the parameter response delay ≤10 minutes, and stabilizes the deviation within 5%, while the average absolute percentage error (MAPE) of prediction is ≤2.1%, which is reduced by 63.8% compared with the gray prediction model, and the deviation rate is ≤3.5% (traditional model 18.7%) under the scene of 10% increase in the proportion of ultra-large ships.

[0045] 2. The operation efficiency optimization effect of the present application is quantified. The present application relies on multi-link dynamic simulation module, dynamic berthing planning module and multimodal transport scheduling module to realize full-process efficiency breakthrough; the multimodal transport scheduling module integrates the information sharing platform, optimizes the path through the ant colony algorithm, shortens the average time of ships in port from 72 hours to 45 hours, reduces by 37.5%, which is much higher than the optimization effect of 15% of the dynamic port capacity model, and the return empty running rate is reduced from 41% to 22%; the multi-link dynamic simulation module couples the calculation of yard operation quantity, promotes the yard turnover rate by 22%, when the yard utilization rate reaches 82%, the turnover time growth elasticity coefficient is reduced from 1.34 to 0.89, effectively alleviating the congestion amplification effect; the dynamic berthing planning module optimizes the 150-dimensional coding berthing scheme based on genetic algorithm, and the average waiting time of ships is shortened by 25%.

[0046] 3. The green development support capability of the present application is prominent. The present application can accurately quantify the comprehensive benefits of green measures by constructing a regression model of shore power utilization rate and throughput through a green simulation module; by optimizing the shore power utilization and new energy ship scheduling scheme, the port can increase the throughput by 12.3% while reducing the carbon emission intensity by 15%, realizing the synergistic optimization of efficiency and environmental protection, and solving the problem that the traditional model cannot quantify the green benefits, resulting in an evaluation deviation of more than 40% of the green investment return rate of some ports, while the present application can accurately predict the throughput variation range and carbon emission reduction amount during the process of increasing the shore power coverage rate from 20% to 50%.

[0047] 4. The emergency response capability of the present application is comprehensively enhanced. The present application improves the extreme scenario response capability based on the uncertainty quantification simulation module and the interference and compensation coupling module; the random event library containing multiple interference factors is established through the uncertainty quantification simulation module, and the throughput prediction results of different confidence intervals are output based on the Monte Carlo algorithm, which improves the simulation credibility of the emergency scenario from less than 60% of the traditional model to more than 85%, and when the port is temporarily closed due to extreme weather, the throughput fluctuation range can be predicted 72 hours in advance, which is beneficial to assist in formulating emergency scheduling schemes; the interference and compensation coupling module quantifies the interruption influence of the TOS / ECS system, combines the artificial intervention coefficient and the artificial replacement efficiency (the ECS efficiency of the fully automated terminal is 0.1-0.3), and through the dynamic integration of the system interruption quantification result and the artificial intervention parameter, the accurate response and operation process adjustment of the system abnormal scenario are realized.

[0048] 5. The decision support value of the present application is significant. The present application supports multi-dimensional decision-making through the parameter matrix construction module and the multi-link dynamic simulation module, for example, in the port expansion planning, the optimal expansion scale is determined as 3 10,000-ton berths through simulating different berth quantity schemes, which reduces the investment by 120 million yuan compared with the initial scheme, and meets the long-term demand of 40% throughput growth, and through subsequent operation verification, the resource utilization rate is improved by 28%, the equipment idle rate is reduced by 15%, and the investment recovery period is shortened by 1.5 years; at the same time, it supports the throughput simulation of different types of terminals (traditional / semi-automatic / fully automatic), and provides data support for intelligent transformation.

[0049] Another aspect of the embodiment of the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the model according to the foregoing.

[0050] Among them, the computer readable medium can be contained in the device, equipment and system of the present disclosure, or can exist independently.

[0051] The computer readable storage medium can be any tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device, and can be electric, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, which more specific examples include but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0052] The computer readable storage medium can also include a data signal carried by a carrier wave or a carrier wave, which carries computer readable program codes, and specific examples include but are not limited to electromagnetic signals, optical signals, or any suitable combination thereof.

[0053] Parameter description: All parameter values involved in this specification are not subjective speculation, but the result of the comprehensive consideration of the safety / efficiency requirements of the application scene, the requirements of industry standards and specifications, and the experience threshold of industry practice. In actual application, the parameters will be fine-tuned according to the relevant scene.

[0054] Finally, it should be pointed out that: the above embodiments are only used to illustrate the technical solutions of the present application, and are not limited thereto; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A port throughput simulation model, characterized in that, include: Database building module: used to integrate historical throughput data, ship AIS dynamic data, equipment operation log data, multimodal transport scheduling data and meteorological statistics to form a basic database; Data preprocessing module: used to detect outliers in the basic database using an improved particle swarm optimization algorithm, fill in missing values ​​using linear interpolation, and standardize data of different dimensions; Parameter matrix construction module: Used to decompose the key influencing factors of the entire port operation process and form a three-dimensional parameter matrix including basic parameters, system interference parameters, artificial compensation parameters, operation impact parameter matrix and random event library parameters; Dynamic parameter self-calibration module: used to build a closed-loop mechanism of "deviation feedback-parameter correction-simulation iteration" based on reinforcement learning, collect ship AIS data and equipment IoT status in real time to dynamically update simulation parameters, and automatically trigger parameter optimization when the deviation between actual throughput and simulated throughput exceeds 5%; Multi-stage dynamic simulation module: It is used to coupled and calculate the workload of each stage of anchorage turnover, wharf front, horizontal transportation, yard operations and gate passage based on the three-dimensional parameter matrix, generate the workload data of each stage of wharf type, generate the throughput of each stage of wharf type by weighted summation of the workload data of each stage of wharf type based on the operation influence parameter matrix, summarize the throughput of each stage of wharf type and generate the total port throughput. Interference and Compensation Coupling Module: Used to quantify the impact of TOS and ECS system outages on the efficiency of various aspects of port operations, and generate compensated operational volume data by combining the human intervention coefficient and the efficiency of human substitution. Uncertainty Quantification Simulation Module: This module is used to construct a random event library containing multiple types of interference factors, train a probability model using historical data, and output throughput prediction results with different confidence intervals based on the Monte Carlo algorithm, thereby realizing uncertainty quantification.

2. The port throughput simulation model according to claim 1, characterized in that, It also includes a dynamic berthing planning module, which is connected to the parameter matrix construction module and the multi-stage dynamic simulation module. The dynamic berthing planning module is used to receive the terminal front operation volume data output by the multi-stage dynamic simulation module, and optimize the 150-dimensional encoded ship berthing scheme based on a genetic algorithm with berth capacity and crane quantity as constraints.

3. The port throughput simulation model according to claim 1, characterized in that, It also includes a multimodal transport scheduling module, which is connected to the database construction module and the interference and compensation coupling module. The multimodal transport scheduling module is used to extract the multimodal transport scheduling data from the database construction module and integrate the information sharing platform to realize the real-time synchronization of port, railway and highway data, plan the optimal transshipment route based on the ant colony algorithm, and output the route optimization results.

4. The port throughput simulation model according to claim 1, characterized in that, It also includes a green simulation module, which is connected to the parameter matrix construction module and the uncertainty quantification simulation module. The green simulation module is used to combine the shore power coverage parameters extracted from the parameter matrix construction module and the anchorage turnover operation data, the wharf front operation data, the horizontal transport operation data, the yard operation data and the gate passage data obtained from the multi-stage dynamic simulation module to construct a regression model of shore power utilization rate and throughput. When shore power coverage increases from 20% to 50%, the magnitude of the change in throughput and the reduction in carbon emissions are quantitatively predicted based on the regression model of shore power utilization and throughput.

5. The port throughput simulation model according to claim 1, characterized in that, The operational impact coefficient matrix includes the impact coefficient of anchorage turnover on throughput. Impact coefficient of dock operations on throughput The impact coefficient of horizontal transport on throughput Impact coefficient of yard operations on throughput The impact coefficient of gate passage on throughput And satisfy .

6. The port throughput simulation model according to claim 1, characterized in that, The operational volume data for each stage of the quay type includes anchorage turnover operational volume data, quayfront operational volume data, horizontal transport operational volume data, yard operational volume data, and gate throughput data; among which: The formula for calculating the anchorage turnover volume data is as follows: ， In the formula, Anchorage capacity (unit: vessels). Average waiting time for ships (unit: hours). The interference coefficient of TOS on the anchorage (value ranges from 0 to 1). TOS system availability (unit: %) Average cargo handling volume per vessel (unit: TEU / vessel). The water diversion efficiency coefficient. For anchorage dispatch operation coefficient, The coefficient represents the human intervention factor. The efficiency of artificial replacement of TOS (x=1, 2, 3, x=1: 0.7-0.9; x=2: 0.5-0.7; x=3: 0.2-0.4). The response time for manual intervention is in hours. Total duration of the statistical period (unit: hours); The formula for calculating the amount of work done at the wharf front is: ， In the formula, Berth operation time (unit: hours). The berth operating time (unit: hours). Number of quay cranes (unit: units). Efficiency of quay crane operations (unit: TEU / hour * unit). The average operating time of the quay crane is in hours. This refers to the dock operation coefficient. ECS system availability (unit: %) This is the TOS interference attenuation coefficient (value 0-1). The TOS dependence is (x=1, 2, 3, x=1: 0.3-0.5, x=2: 0.6-0.8, x=3: 0.8-1.0). This is the ECS interference attenuation coefficient (values ​​range from 0 to 1, with 0 when x=1). ECS dependency (x=1, 2, 3, x=1: 0, x=2: 0.5-0.7, x=3: 0.9-1.0). Efficiency of manual replacement of ECS (x=1, 2, 3, x=1: 0, x=2: 0.4-0.6, x=3: 0.1-0.3); The formula for calculating horizontal transport volume data is: ， In the formula, The number of horizontal transport equipment (unit: units). Average operating efficiency of the equipment (unit: TEU / hour * unit). Average daily operating time of the equipment (unit: hours). The smoothness coefficient of the transportation route. This is the ECS interference path attenuation coefficient (value 0-1). ECS control dependence (x=1, 2, 3, x=1: 0, x=2: 0.4-0.6, x=3: 0.7-1.0). This refers to the horizontal transport operation coefficient. The formula for calculating yard operation volume data is: ， In the formula, The amount of cargo stored in the yard (unit: TEU). Total cargo volume stored in the yard (unit: TEU). Number of yard bridges (unit: units). Efficiency of yard crane operations (unit: TEU / hour * unit). The average working time for the yard bridge is in hours. This refers to the yard operation coefficient. The disturbance coefficient of TOS to the stockyard (values ​​range from 0 to 1). The TOS plan dependency ratio is (x=1, 2, 3, x=1: 0.2-0.4, x=2: 0.5-0.7, x=3: 0.8-1.0). This is the interference coefficient of ECS on the storage yard (values ​​range from 0 to 1, with 0 when x=1). ECS field bridge dependence (x=1, 2, 3, x=1: 0, x=2: 0.3-0.5, x=3: 0.8-1.0); The formula for calculating the throughput data at the gate is: ， In the formula, The gate throughput efficiency (unit: vehicles / hour). Number of gate channels (unit: channels). The time for passage through the gate (unit: hours). Average container truck capacity (unit: TEU / vehicle). This is the gate operation coefficient. The interference coefficient of TOS on the gate (value ranges from 0 to 1). The dependence of TOS on the gate (x=1, 2, 3, 0.3-0.5 when x=1, 0.6-0.8 when x=2, and 0.8-1.0 when x=3).

7. The port throughput simulation model according to claim 6, characterized in that, The formula for calculating the throughput of the sub-terminal type is as follows: ， In the formula, Let x be the total throughput of terminal x (including interference and compensation), where x=1,2,3 correspond to traditional terminal, semi-automated terminal, and fully automated terminal, respectively. This refers to the anchorage turnover data for wharf type x. This refers to the data on the amount of work being done at the wharf front for wharf type x. This is the horizontal transport volume data for terminal type x. This refers to the yard operation volume data for terminal type x. This refers to the throughput data at the gate of terminal x. The formula for calculating the total throughput of the port is as follows: In the formula, For the throughput of traditional ports, For the throughput of semi-automated terminals, This represents the throughput of a fully automated terminal.

8. The port throughput simulation model according to claim 1, characterized in that, It also includes a model verification and calibration module, which is connected to the multi-stage dynamic simulation module and the dynamic parameter self-calibration module respectively, and is used to verify and optimize the model accuracy, including: Historical backtesting: Using historical port operation data, the output results of the multi-stage dynamic simulation module are verified to ensure that the RMSE value is <8.6% and the MAPE value is ≤2.1%. Sensitivity analysis: Identify key parameters, establish a parameter adjustment threshold system, and trigger the dynamic parameter self-calibration module when the threshold is exceeded; Field test: The system was tested at a coastal port for 3 months, collecting data from 120 ships to calibrate the model parameters.

9. An electronic device, characterized in that, include: One or more processors; A storage unit for storing one or more programs, which, when executed by one or more processors, enable the one or more processors to implement the port throughput simulation model according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it can realize the port throughput simulation model according to any one of claims 1 to 8.

Images