Urban rail transit hub transfer facility optimal configuration method and system

Abstract

The invention discloses an optimal configuration method and system for transfer facilities of an urban rail transit hub. The method comprises the following steps: constructing an emergency and a trigger condition thereof; collecting urban rail transit transportation data; designing an emergency and passenger flow data transfer function, and calculating an influence value of the emergency on the passenger flow data; constructing a basic passenger flow prediction model, embedding an emergency triggering model, and newly adding an event embedding layer containing an emergency and a passenger flow data transfer function to generate a passenger flow prediction model; utilizing the passenger flow prediction model to predict passenger flow data of each station; and by applying a multi-objective optimization algorithm and combining the predicted passenger flow data, obtaining a configuration strategy of an optimal transfer facility and adjusting transfer facility configuration, monitoring transfer efficiency and continuously adjusting parameters of the multi-objective optimization algorithm or parameters of the passenger flow prediction model through a feedback mechanism until the preset transfer efficiency is reached. According to the method, the transfer facility configuration can be optimized and the transfer efficiency can be improved for the complex and changeable transportation conditions of the urban rail transit.

Description

Technical Field

[0002] The present application relates to the technical field of urban rail transit hub configuration optimization, and specifically to a method and system for optimizing the configuration of transfer facilities in an urban rail transit hub. Background Art

[0003] In urban rail transit systems, an efficient and rational layout of transfer facilities is key to improving overall system efficiency. Currently, with the acceleration of urbanization and the growth of urban populations, urban rail transit passenger volume has increased significantly, placing higher demands on transfer efficiency and passenger experience at rail transit hubs. However, the design of transfer facilities in existing urban rail transit hubs relies primarily on traditional planning theories and experience, lacking accurate passenger flow forecasting and dynamic optimization algorithms. This, to a certain extent, limits further improvements in transfer efficiency.

[0004] The optimization of existing transfer facilities is primarily achieved through static analysis, with preliminary planning based on historical data, followed by adjustments based on actual operational conditions. Specifically, planners collect historical passenger flow data, including passenger volume and transfer ratios over different time periods, and use this historical data to establish a preliminary planning model to determine the approximate scale and layout of the transfer facilities. After the rail transit hub is completed and put into operation, the transfer facilities are then partially adjusted based on actual passenger flow changes and operational conditions. Other methods rely on experienced professionals to design and configure based on the experience of similar projects, but this approach lacks systematicity and scientificity.

[0005] These methods often fail to adapt in real time to the increasingly complex passenger flow demands and rapidly changing urban traffic patterns. Traditional static analysis methods based on historical data cannot accurately predict changes in passenger flow under unexpected circumstances, leading to frequent transfer bottlenecks. When faced with special events or emergencies, such as large-scale events or inclement weather, passenger flow can suddenly change, and existing optimization configuration methods cannot respond in a timely manner, resulting in long wait times for passengers and inconvenience in transfers. Furthermore, traditional planning theories lack flexibility and adaptability when faced with complex and changing transportation network structures, making it difficult to meet the development needs of modern urban rail transit. Summary of the Invention

[0006] In order to cope with the complex and changeable transportation conditions of urban rail transit, adaptively optimize the configuration of transfer facilities, and improve passenger transfer efficiency, the present application provides a method and system for optimizing the configuration of transfer facilities in urban rail transit hubs.

[0007] In a first aspect, the present application provides a method for optimizing the configuration of transfer facilities in an urban rail transit hub, comprising: Considering the influencing factors of emergencies in urban rail transit hub transfers, construct emergencies and their triggering conditions; collect real-time passenger flow data, emergency triggering condition correlation data, and transfer facility configuration data of each station in urban rail transit; combine the collected historical real-time passenger flow data and emergency triggering condition correlation data in urban rail transit to design a transfer function between emergencies and passenger flow data, and calculate the impact value of emergencies on passenger flow data; based on the deep learning algorithm, construct a basic passenger flow prediction model, embed the emergency trigger model in the basic passenger flow prediction model to achieve adaptive adjustment of the prediction parameters in the basic passenger flow prediction model, and simultaneously add an event embedding layer containing emergency triggering conditions, emergency and passenger flow data transfer function to the basic passenger flow prediction model to obtain the impact value of emergencies on passenger flow data and input it into the passenger flow prediction model to generate a passenger flow prediction model; use the passenger flow prediction model to predict passenger flow data at each station; For each station, a multi-objective optimization algorithm is used, combined with predicted passenger flow data, to obtain the optimal transfer facility configuration strategy, minimizing the sum of queue waiting costs and transfer facility operating resource costs over multiple time periods while meeting transfer facility capacity constraints. For each station, the transfer facility configuration is adjusted according to the optimal transfer facility configuration strategy, and the transfer efficiency of the station is monitored in real time. The multi-objective optimization algorithm parameters or passenger flow prediction model parameters are continuously adjusted through the feedback mechanism until the transfer efficiency reaches the preset transfer efficiency.

[0008] By adopting the above solution, we construct emergency events and their triggering conditions, design transfer functions based on historical data to calculate the impact value, and then use deep learning algorithms to build and improve passenger flow prediction models to predict passenger flow data. Then, we use multi-objective optimization algorithms to obtain the optimal configuration strategy. Finally, we adjust transfer facilities according to the strategy and monitor and adjust algorithm parameters in real time. This realizes the intelligent optimization configuration of transfer facilities, significantly improves transfer efficiency, and reduces passenger waiting time.

[0009] Preferably, it also includes: Collect building data around each station in urban rail transit, and use spatial analysis technology to extract GIS spatial characteristics of each station, including land use type and road network characteristics; the land use type includes commercial land and residential land; the road network characteristics include high road network density, medium road network density and low road network density; According to different GIS spatial feature combinations, different transfer facility capacity configuration ratio weight ranges matching each GIS spatial feature combination are pre-designed, so that the weight of the elevator or escalator transfer facility capacity configuration ratio in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing commercial land types is higher than that matching the spatial feature combination containing residential land types, and the transfer facility type combination with a smaller average distance in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing high road network density is higher than that matching the spatial feature combination containing medium road network density, and the transfer facility type combination with a smaller average distance in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing medium road network density is higher than that matching the spatial feature combination containing low road network density, and the transfer facility type combination with a smaller average distance in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing medium road network density is higher than that matching the spatial feature combination containing low road network density. In the process of using a multi-objective optimization algorithm and combining predicted passenger flow data to obtain the optimal strategy for optimizing the configuration of transfer facilities, the weight range of the capacity configuration proportion of different transfer facilities in the configuration optimization strategy of transfer facilities is restricted to meet the matching weight range of the capacity configuration proportion of different transfer facilities.

[0010] By adopting the above scheme, the impact of factors such as land use type and road network layout around different stations on transfer demand is also considered. Surrounding building data is collected and GIS spatial features are extracted. The weight range of transfer facility capacity configuration that matches different GIS spatial feature combinations is pre-designed, and restrictions are imposed in the process of obtaining the optimal configuration strategy. This enhances the model's adaptability to complex transportation networks, making the optimized configuration of transfer facilities more flexible and better matching the needs of urban development.

[0011] Preferably, the real-time passenger flow data in historical urban rail transit and the associated data on emergency trigger conditions are combined to design a transfer function between emergencies and passenger flow data, and the formula for calculating the impact of emergencies on passenger flow data is: ΔQ(t)=∑[f i (E i ,S i ,D i ,R i )*Q Base (t)] Where ΔQ(t) is the impact value of passenger flow data caused by the emergency at time t; f i (E i ,S i ,D i ,R i ) is the transfer function for the i-th sudden event, which depends on the event type E i , severity S i Duration Di And the impact range R i ; Among them, emergency type E i Including transfer facility failure at this station, facility failure at adjacent stations, suspension or addition of trunk stations, extreme weather or large-scale events and holidays; severity level S i Set to a value of 1-10 to indicate the urgency and severity of the emergency; duration D i It refers to the length of time from the beginning to the end of an emergency in minutes, hours or days; the impact range R i Refers to the scope of a single or several sites; Q Base (t) is the predicted passenger flow base value at time t when no emergency occurs.

[0012] By adopting the above solution, the calculation formula for the impact value of emergencies on passenger flow data is clarified from the perspectives of emergency type, severity, duration and impact range, and the impact of emergencies on passenger flow data is accurately calculated, thereby ensuring more accurate prediction of passenger flow data for each station in the future.

[0013] Preferably, for each station, a multi-objective optimization algorithm is used, combined with predicted passenger flow data, to obtain a configuration optimization strategy for transfer facilities, thereby minimizing the sum of queue length and resource cost in multiple time periods while satisfying transfer facility capacity constraints and resource utilization constraints. The multi-objective optimization algorithm formula is: The objective function is: Where W q (t) represents the waiting time of the queue in the t-th time period. The specific time period is obtained by artificial division based on a 24-hour day, and it is assumed that the transfer facility service process conforms to the queuing theory model and is determined according to the transfer flow and transfer facility configuration strategy; C(t) represents the transfer facility operation resource cost in the t-th time period, which is determined according to the transfer facility configuration strategy and the transfer facility operation unit cost; W queue and W cost Both represent weight coefficients; The constraints are: Transfer facility capacity ≥ Q predict (t) Where, the transfer facility capacity is obtained based on the configuration of the transfer facility and the time period t; Q predict (t) is the transfer traffic demand predicted in the tth time period; The decision variable is the configuration strategy P of the transfer facilities, which can be expressed as: P(K,M (K) , X (K) , V (K) ) K={k1,k2,……k n} M (ki) ≤M (ki) max V (ki) ≤V (ki) max In the formula, the configuration strategy P of transfer facilities depends on the type of transfer facilities K and the number of transfer facilities M under the condition of the triggering emergency event. (K) , the location X of the corresponding transfer facility under the condition of transfer facility selection K (K) , the operating speed V of the corresponding transfer facility under the condition of transfer facility selection K (K) ;k i represents the i-th type of transfer facility, i is 1...n; M (ki) max represents the maximum number of the i-th transfer facilities at the current station; V (ki) max represents the maximum safe operating rate of the i-th transfer facility at the current station.

[0014] By adopting the above scheme, the objective function, constraints and decision variables of the multi-objective optimization algorithm are clarified. The algorithm can be used in combination with the predicted passenger flow data to obtain the configuration optimization strategy of transfer facilities, thereby minimizing the sum of transfer waiting time and resource costs in multiple time periods, improving transfer efficiency while controlling costs.

[0015] Preferably, the configuration strategy range of the transfer facilities corresponding to the decision variables is dynamically adjusted according to the impact assessment level of the triggered emergency event; specifically, the range includes: Determine the impact assessment level of the triggered emergency event based on the passenger flow impact value range of the passenger flow data caused by the emergency event. Each passenger flow impact value range corresponds to an impact assessment level of a type of emergency event. The impact assessment levels include major impact and minor impact. In response to emergencies triggered by an impact assessment level of significant impact, the transfer facility types in the transfer facility configuration strategy are adjusted to add emergency transfer facility type options. The location, maximum number, and maximum safe operating rate of the transfer facilities corresponding to the newly added emergency transfer facility types are supplemented.

[0016] By adopting the above scheme, the transfer facility configuration strategy range is dynamically adjusted according to the emergency impact assessment level, making the transfer facility configuration more flexible to respond to emergencies of different levels, ensuring that passenger flow demand is met by adding new types of emergency transfer facilities when major impact events occur, further improving transfer efficiency and reducing passenger waiting time.

[0017] Preferably, it also includes: For each station, consider the service life of different types of transfer facilities at each station and obtain the capacity depreciation coefficient of different types of transfer facilities within different service life; Introducing the capacity depreciation coefficient of the transfer facilities into the constraints, the constraints are readjusted to: ∑ψ i * Under the condition of P, the capacity of the i-th type transfer facility in time period t ≥ Q predict (t) Where, ψ i Represents the capacity loss coefficient of the i-th type transfer facility.

[0018] By adopting the above scheme, the capacity depreciation coefficient is calculated considering the age of the transfer facilities and constraints are introduced, so that the transfer facility configuration strategy is more in line with the actual situation, avoiding the capacity error caused by facility aging, and improving the accuracy of the strategy.

[0019] Preferably, it also includes: For each station, a digital twin 3D model of the station including the transfer facilities of the urban rail transit hub is constructed; Under the conditions of setting the predicted transfer flow demand and the configuration strategy of the optimal transfer facilities, the transfer simulation is completed using simulation technology to obtain the transfer efficiency of the transfer simulation process; After determining that the transfer efficiency obtained in the transfer simulation process is lower than the preset transfer efficiency, the multi-objective optimization algorithm parameters are continuously adjusted through the feedback mechanism until the transfer efficiency obtained in the transfer simulation process is higher than the preset transfer efficiency, and the configuration strategy of the optimal transfer facility is re-obtained to replace the configuration strategy adjustment of the optimal transfer facility originally obtained.

[0020] By adopting the above solution, a digital twin three-dimensional model is constructed and transfer simulation is performed to evaluate transfer efficiency. When the simulation efficiency is lower than the preset value, the multi-objective optimization algorithm parameters are adjusted through the feedback mechanism and the configuration strategy is re-acquired, thus achieving dynamic optimization of the transfer facility configuration and improving transfer efficiency and passenger experience.

[0021] In a second aspect, the present application provides a system for optimizing the configuration of transfer facilities in an urban rail transit hub, comprising: Rail transit emergency event construction module, which is used to consider the influencing factors of emergency events at urban rail transit hub transfers and construct emergency events and their triggering conditions; Rail transit data collection module, used to collect real-time passenger flow data, emergency trigger condition correlation data, and transfer facility configuration data at each station in urban rail transit; The rail transit passenger flow data prediction module is used to combine the collected real-time passenger flow data of historical urban rail transit and the associated data of emergency trigger conditions, design the transfer function between emergency events and passenger flow data, and calculate the impact value of emergency events on passenger flow data; based on the deep learning algorithm, a basic passenger flow prediction model is constructed, and the emergency trigger model is embedded in the basic passenger flow prediction model to achieve adaptive adjustment of the prediction parameters in the basic passenger flow prediction model. At the same time, an event embedding layer containing emergency trigger conditions, emergency events and passenger flow data transfer function is added to the basic passenger flow prediction model to obtain the impact value of emergency events on passenger flow data and input it into the passenger flow prediction model to generate a passenger flow prediction model; the passenger flow prediction model is used to predict the passenger flow data of each station; The rail transit transfer facility configuration acquisition module is used to apply a multi-objective optimization algorithm to each station, combined with predicted passenger flow data, to obtain the optimal transfer facility configuration strategy, thereby minimizing the sum of queue waiting costs and transfer facility operating resource costs over multiple time periods while meeting transfer facility capacity constraints. The rail transit transfer facility configuration adjustment module is used to adjust the transfer facility configuration for each station according to the optimal transfer facility configuration strategy, monitor the transfer efficiency of the station in real time, and continuously adjust the multi-objective optimization algorithm parameters or passenger flow prediction model parameters through the feedback mechanism until the transfer efficiency reaches the preset transfer efficiency.

[0022] By adopting the above scheme, the influencing factors of emergencies are taken into account and relevant conditions are constructed. Multiple types of data are collected, and the transfer function is designed in combination with historical data to calculate the impact value. The deep learning algorithm is used to construct a passenger flow prediction model containing an emergency trigger model and an event embedding layer to predict passenger flow. The multi-objective optimization algorithm is used to obtain the optimal transfer facility configuration strategy, which can realize the intelligent optimization configuration of transfer facilities, minimize the sum of queue waiting time and operating resource costs, and adjust the algorithm parameters through the feedback mechanism to make the transfer efficiency higher than the preset value.

[0023] In a third aspect, the present application provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the method as described above.

[0024] In a fourth aspect, the present application provides a computer device, which includes a memory, a processor, and a program stored and executable on the memory, and the program implements the steps of the above method when executed by the processor.

[0025] In summary, this application has the following beneficial effects: 1. Construct emergency events and their triggering conditions, use deep learning algorithms to build a passenger flow prediction model and embed it into the emergency event triggering model. Combined with a multi-objective optimization algorithm, the optimal transfer facility configuration strategy can be obtained based on the predicted passenger flow data, realizing intelligent optimization of transfer facilities, significantly improving transfer efficiency, and reducing passenger waiting time. 2. By integrating real-time passenger flow data, data related to emergency trigger conditions, and GIS spatial features, we pre-designed a range of transfer facility capacity allocation weights that matches different GIS spatial feature combinations and imposes constraints in the process of obtaining the optimal allocation strategy, thus enhancing the model's adaptability to complex transportation networks and emergencies. 3. Accurately designing the layout and size of transfer facilities using passenger flow prediction models and multi-objective optimization algorithms, combined with digital twin 3D models and simulation technology, improves the transfer experience for passengers and reduces maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a flow chart of the method for optimizing the configuration of transfer facilities in an urban rail transit hub according to a specific embodiment; Figure 2 It is a structural diagram of the urban rail transit hub transfer facility optimization configuration system described in a specific embodiment. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0028] like Figure 1 As shown, the embodiment of the present application discloses a method for optimizing the configuration of transfer facilities in an urban rail transit hub, including the steps of data collection, passenger flow prediction, configuration strategy acquisition, and facility adjustment and monitoring. Specifically, by collecting various types of data and combining deep learning and multi-objective optimization algorithms, dynamic optimization of transfer facilities is achieved, thereby improving transfer efficiency, reducing costs, and meeting passenger flow needs. The specific steps are as follows: S1. Collect transportation data of each urban rail transit station.

[0029] Specifically, to predict passenger flow data at each urban rail transit station, we consider the impact factors of sudden incidents at urban rail transit hub transfers, analyze and construct sudden incidents and their triggering conditions. These factors can be multifaceted, such as station facility failures, failures at adjacent stations, closures or additions of trunk line stations, extreme weather conditions, major events, and holidays.

[0030] In this embodiment, the constructed emergencies include: facility failure at this site, failure at an adjacent site, suspension or addition of a trunk site, extreme weather or large-scale events and holidays; the trigger conditions corresponding to the need to integrate information from the meteorological department, event organizer, multi-site equipment monitoring system, etc. include: receiving a facility failure signal at this site uploaded by the equipment monitoring system of this site, receiving a facility failure signal at an adjacent site uploaded by the equipment monitoring system of an adjacent site, receiving suspension or addition information of a site uploaded by the trunk equipment monitoring system, receiving extreme weather warning information issued by a meteorological station, or receiving event information issued by an event organizer and holiday information issued by this site, etc.

[0031] On the basis of constructing emergencies and their corresponding triggering conditions, urban rail transit transportation data are collected, including: real-time passenger flow data of each station in urban rail transit transportation, including the number of arrivals, departures and transfers, as well as the passenger flow density in different time periods, etc., transfer facility configuration data of each station in urban rail transit transportation, including the type and number of transfer facilities, structural data and operation data of each transfer facility, etc., and emergency event trigger condition association data of each station in urban rail transit transportation, including: facility failure information of this station uploaded by the equipment monitoring system of this station for the current and future periods, received facility failure information of adjacent stations uploaded by the information monitoring system of adjacent stations for the current and future periods, received site suspension or new addition information set for the current and future periods uploaded by the trunk equipment monitoring system, received extreme weather warning information issued by the meteorological station for the current and future periods, or event information issued by the event organizer for the current and future periods, and holiday information released by this station, etc.

[0032] S2. Based on the collected transportation data of each urban rail transit station, the passenger flow data of each station is predicted.

[0033] Considering that the existing technology directly uses deep learning algorithms to build passenger flow prediction models, it is impossible to accurately predict passenger flow changes in the face of various emergencies, and thus it is impossible to adjust the transfer facility layout and optimize the algorithm parameters in a timely manner, resulting in low transfer efficiency. To avoid the above situation, it is chosen to embed an event trigger module in the traditional basic passenger flow prediction model and design an event-passenger flow transfer function to further optimize the passenger flow prediction results.

[0034] First, combining the real-time passenger flow data of historical urban rail transit and the associated data of emergency trigger conditions, we designed a transfer function between emergency events and passenger flow data, and calculated the impact of emergencies on passenger flow data. The formula is: ΔQ(t)=∑[f i (E i ,S i ,Di ,R i )*Q Base (t)] Where ΔQ(t) is the impact value of passenger flow data caused by the emergency at time t; f i (E i ,S i ,D i ,R i ) is the transfer function for the i-th sudden event, which depends on the event type E i , severity S i Duration D i And the impact range R i ; For the i-th emergency event E i The transfer function can be specified as: f i (E i ,S i ,D i ,R i )=α*S i +β*D i +γ*R i Where α, β, and γ are weight coefficients obtained through regression analysis of historical data on specific types of emergencies. They reflect the different degrees of influence of severity, duration, and impact range on passenger flow changes. The weight coefficients obtained through regression analysis of historical data on different types of emergencies are different.

[0035] Among them, emergency type E i Including transfer facility failure at this station, facility failure at adjacent stations, suspension or addition of trunk stations, extreme weather or large-scale events and holidays; severity level S i Set to a value of 1-10 to indicate the urgency and harm of the emergency, and the duration D i It refers to the length of time from the beginning to the end of an emergency in minutes, hours or days; the impact range R i Refers to the scope of a single or several sites; Q Base (t) is the predicted passenger flow base value at time t, assuming no emergencies occur. Specifically, the severity, duration, and impact range of different types of emergencies are included in the received fault information, suspension or new service information, weather information, major events, and holiday information, and are determined accordingly.

[0036] Secondly, a basic passenger flow prediction model is constructed based on a deep learning algorithm. An event embedding layer is added to the basic passenger flow prediction model, which includes emergency trigger conditions and a transfer function between emergency events and passenger flow data. The embedding layer determines whether to trigger an emergency based on the input emergency trigger condition association data. The impact value of the emergency on the passenger flow data is calculated in combination with the passenger flow data transfer function. This impact value of the emergency on the passenger flow data is used as input and is used together with the input passenger flow data to predict the output passenger flow data. In addition, in addition to ensuring the accuracy of passenger flow data prediction from the input perspective, the emergency event trigger model is embedded in the basic passenger flow prediction model to achieve adaptive adjustment of the prediction parameters in the basic passenger flow prediction model. Specifically, the embedded emergency event trigger model uses unique hot encoding to represent the event type. When the event trigger module recognizes the occurrence of a specific emergency event, it will automatically adjust the parameters of the passenger flow prediction model. The parameter adjustment is based on pre-set rules and strategies (such as: increasing the weight of the corresponding type of emergency event trigger condition associated data) to ensure that the model can accurately predict the passenger flow change characteristics under different event modes. For example, if there is a large-scale event, the prediction weight of the large-scale event trigger condition associated data of the large-scale event radiation site is increased.

[0037] Finally, a passenger flow prediction model is generated by embedding an event trigger module in the basic passenger flow prediction model and integrating the emergency trigger conditions and the emergency-passenger flow transfer function into the basic passenger flow prediction model; the generated passenger flow prediction model is used to predict the passenger flow data of each station, and the entry, exit and transfer volumes of each station as well as the passenger flow density in different time periods are obtained. Specifically, for a station, the passenger flow data of the current period is collected, and the emergency trigger condition association data of the next period received is input into the passenger flow prediction model to predict and obtain the passenger flow data of the corresponding station in the next period.

[0038] S3. Use multi-objective optimization algorithms and combine them with predicted passenger flow data to obtain the optimal transfer facility configuration strategy.

[0039] Specifically, considering the differences in passenger flow data predicted for each station and the different configurations of transfer facilities at each station, a multi-objective optimization algorithm is used to obtain the optimal transfer facility configuration strategy for each station. Taking each station as an example, the optimal transfer facility configuration strategy includes: First, considering that the ultimate goal is to improve efficiency and reduce the waiting time of passengers in front of transfer facilities, and taking a 24-hour day as an example, it can be artificially stipulated, such as: according to day and night, the daytime is divided into a time period t of 1 minute to 10 minutes, and the night is divided into a time period of 10 minutes to 30 minutes. In this embodiment, the objective function is to minimize the sum of the queue waiting cost and the transfer facility operation resource cost in multiple time periods, where the multiple time periods T = Nt, N is a positive integer, and the specific value of N can be artificially set according to the needs of the station. The corresponding objective function formula is: Where W q (t) represents the waiting time of the queue in the tth time period, C(t) represents the operating resource cost of the transfer facility in the tth time period, which is determined according to the transfer facility configuration strategy and the unit cost of the transfer facility operation; W queue and W cost Both represent weight coefficients; Specifically, it is assumed that the service process of transfer facilities conforms to the queuing theory model (such as the M / M / c queuing system, i.e., Poisson arrival, exponential service, and multiple service stations), the transfer flow Q is uniformly distributed within the time period t, and the passenger flow arrivals of each transfer facility are independent of each other. The operating service time of different transfer facilities follows an exponential distribution, and the parameter W is related to the operating rate; q (t) is determined based on the transfer flow and transfer facility configuration strategy. The formula is: Assume that the transfer flow Q is distributed to each facility according to the attraction of the transfer facility location X(K) and the transfer facility type K (e.g., k1 represents escalator type, k2 represents elevator type, etc.), and the attraction weight is w (ki) The formula is: Where, For passengers to transfer facilities i The average walking distance of the position X(k i ) determine; η(k i ) represents the transfer facility type k i The efficiency coefficient is pre-set (e.g. 1 for escalator, 0.8 for elevator, etc.); then type k i The passenger arrival rate of the transfer facility in time period t is: According to the M / M / c queuing model, the waiting time calculation for a single type of transfer facility includes: For type k i The average waiting time for transfer facilities is W q(ki) It can be calculated according to the queuing theory formula: Where u (ki) Indicates type k i The single-unit service rate of the transfer facility, i.e. (operating rate); Indicates type k i The number of transfer facilities is determined based on the transfer facility configuration strategy. P0 is the probability that all transfer facilities are idle, as shown in the following formula: Considering all transfer facility types K, the total average waiting time is the weighted average of the waiting time of each transfer facility, where the weight is the proportion of passenger flow carried by the transfer facility. The average waiting time W for all types of transfer facilities is calculated as follows: q : The above formula can be used to calculate the queue waiting time for passengers in time period t based on the decision variable, namely the selected transfer facility configuration strategy. Similarly, the transfer facility operating resource cost C(t) for time period t can be calculated based on the selected transfer facility configuration strategy and the unit operating costs of different transfer facilities.

[0040] Secondly, for selecting different transfer facility configuration strategies while satisfying the transfer facility capacity constraint, the constraint formula is: Transfer facility capacity ≥ Q predict (t) The transfer facility capacity is obtained based on the configuration of the transfer facility and the time period t, including different transfer facility types and the capacity of a single operation of the corresponding type of transfer facility (passenger capacity, such as the elevator capacity of 15 people), the transfer facility operation rate, the probability of the transfer facility system being idle, and the time period t; Q predict (t) is the predicted transfer flow demand in the tth time period. Considering that transfers at some stations overlap with some exit routes, and some exiting passengers will also use transfer facilities, the corresponding transfer flow demand is the transfer volume plus the exit volume of the overlapping part of the station exit route and the transfer route.

[0041] Among them, the decision variable, that is, the decision variable is the configuration strategy P of the transfer facilities, which is expressed as: P(K,M (K) , X (K) , V (K) ) K={k1,k2,……k n} In the formula, the configuration strategy P of transfer facilities depends on the type of transfer facilities K and the number of transfer facilities M under the condition of the triggering emergency event.(K) , the location X of the corresponding transfer facility under the condition of transfer facility selection K (K) , the operating speed V of the corresponding transfer facility under the condition of transfer facility selection K (K) ;k i represents the i-th type of transfer facility, where i ranges from 1 to n; Indicates the maximum number of the i-th transfer facilities at the current station; It shows the maximum safe operating rate of the i-th transfer facility at the current station.

[0042] According to the above objective function and its corresponding constraints, we choose to use genetic algorithm or particle swarm optimization algorithm to solve the optimal transfer facility configuration strategy P * .

[0043] In addition, considering that different emergency events have different impact levels, the corresponding transfer facilities that can be authorized for use may vary. For example, emergency events with a significant impact often require additional emergency transfer facilities. To further ensure the optimal transfer facility configuration strategy, the method further includes: The range of decision variables that can be selected, that is, the configuration strategy range of transfer facilities is dynamically adjusted according to the impact assessment level of the triggered emergency event; specifically, the impact assessment level of the triggered emergency event is determined according to the passenger flow quantity impact value range interval in which the impact value of the passenger flow data caused by the calculated emergency event is located, and each passenger flow quantity impact value range interval corresponds to the impact assessment level of a type of emergency event; the impact assessment level includes a major impact set corresponding to the first passenger flow quantity impact value range interval, and a non-major impact set corresponding to the second passenger flow quantity impact value range interval; for the emergency event triggered with an impact assessment level of major impact, the transfer facility type in the configuration strategy of the transfer facility is adjusted to add an emergency transfer facility type option (such as n-n+m), and the location, maximum number and maximum safe operating rate of the transfer facilities corresponding to the newly added emergency transfer facility types are supplemented.

[0044] In addition, considering the impact of the age of transfer facilities on the capacity of transfer facilities, the method further includes: for each station, considering the age of different types of transfer facilities at each station, querying and obtaining the capacity depreciation coefficients of different types of transfer facilities within different service years according to the instructions of different transfer facilities; introducing the capacity depreciation coefficients of the transfer facilities into the constraint conditions, and readjusting the constraint conditions to: Where, ψ i Represents the capacity loss coefficient of the i-th type transfer facility.

[0045] S4. Adjust the transfer facility configuration according to the optimal transfer facility configuration strategy, monitor the transfer efficiency of the station in real time, and continuously adjust the multi-objective optimization algorithm parameters or passenger flow prediction model parameters through the feedback mechanism.

[0046] Specifically, for each station, the transfer facility configuration is adjusted according to the optimal transfer facility configuration strategy, that is, the transfer facility configuration is adjusted according to the optimal transfer facility configuration strategy within the predicted time period t, thereby improving the passenger transfer efficiency of each station; The transfer efficiency of each station is monitored in real time, and the multi-objective optimization algorithm parameters or the passenger flow prediction model parameters are continuously adjusted through the feedback mechanism. That is, if the actual transfer efficiency within the prediction period t does not reach the preset transfer efficiency, it indicates that the current transfer facility configuration cannot well meet the passenger transfer needs. The possible reason is that there is a problem with the predicted passenger flow, or the configuration of the solved optimal transfer facility is inaccurate. The actual passenger flow data collected within the prediction period t can be compared with the predicted passenger flow data. If the prediction error is less than the preset passenger flow error, the multi-objective optimization algorithm parameters are adjusted accordingly. Otherwise, the passenger flow prediction model parameters are adjusted until the transfer efficiency monitored within the preset period reaches the preset transfer efficiency; the preset period is determined manually.

[0047] In a specific embodiment, the transfer facility configuration strategy is further refined by considering the geographic information surrounding the station. Different land use types and road network characteristics will affect passenger flow distribution and transfer demand. Setting different weight ranges based on these characteristics can make the configuration of transfer facilities more accurately meet actual needs, improve the adaptability and rationality of the system, and thus further enhance transfer efficiency and passenger experience. The method also includes: Collect building data around each station in urban rail transit, and use spatial analysis technology to extract GIS spatial characteristics of each station, including land use type and road network characteristics; the land use type includes commercial land and residential land; the road network characteristics include high road network density, medium road network density and low road network density; According to different GIS spatial feature combinations, different transfer facility capacity configuration ratio weight ranges matching each GIS spatial feature combination are pre-designed to meet the following requirements: the spatial feature combination containing commercial land types has a higher weight for the different transfer facility capacity configuration ratios than the spatial feature combination containing residential land types, because commercial areas usually have a larger passenger flow and a higher demand for vertical transportation; and the transfer facility type combination with a smaller average distance has a higher weight for the different transfer facility capacity configuration ratios than the spatial feature combination containing medium road network density. Facility type combinations have higher capacity weights. This is because the spatial characteristics of high road network density correspond to smaller transfer facility spacing requirements than those of medium road network density. Transfer facility combinations with smaller average distances are more likely to be chosen by passengers, and accordingly, transfer facility combinations with smaller average distances have a larger capacity weight than other transfer facility combinations. Similarly, transfer facility type combinations with smaller average distances have higher capacity weights than other transfer facility type combinations in the weights of different transfer facility capacity configurations that match spatial characteristics combinations with medium road network density compared to those with low road network density. The specific ranges of different transfer facility capacity configuration weights that match each GIS spatial characteristic combination are trained and generated using a deep learning model using different GIS spatial characteristic combinations to determine the ranges of different transfer facility capacity configuration weights that meet the conditions.

[0048] In the process of using a multi-objective optimization algorithm and combining predicted passenger flow data to obtain the optimal strategy for optimizing the configuration of transfer facilities, the weight range of the capacity configuration proportion of different transfer facilities in the configuration optimization strategy of transfer facilities is restricted to meet the matching weight range of the capacity configuration proportion of different transfer facilities.

[0049] In a specific embodiment, a platform for verifying and optimizing transfer facility configuration strategies is provided by combining a digital twin 3D model with simulation technology. Prior to actual implementation, simulations are conducted in a virtual model to identify potential problems in advance and adjust configuration strategies in a timely manner. This avoids low transfer efficiency in actual applications, improves the accuracy and reliability of decision-making, and thereby better meets passenger flow demand and enhances transfer efficiency. The method further includes: For each station, a digital twin 3D model of the station including the transfer facilities of the urban rail transit hub is constructed; Under the conditions of setting the predicted transfer flow demand and the configuration strategy of the optimal transfer facilities, the transfer simulation is completed using simulation technology to obtain the transfer efficiency of the transfer simulation process; After determining that the transfer efficiency of the transfer simulation process is lower than the preset transfer efficiency, the multi-objective optimization algorithm parameters are continuously adjusted through the feedback mechanism until the transfer efficiency of the transfer simulation process is higher than the preset transfer efficiency. The adjusted multi-objective optimization algorithm parameters are used to re-obtain the optimal transfer facility configuration strategy to replace the originally obtained optimal transfer facility configuration strategy adjustment.

[0050] like Figure 2 As shown, the embodiment of the present application discloses a system for optimizing the configuration of transfer facilities in an urban rail transit hub, specifically comprising: Rail transit emergency event construction module 101, used to consider the influencing factors of emergency events at urban rail transit hub transfers, and construct emergency events and their triggering conditions; Rail transit data collection module 102, used to collect real-time passenger flow data, emergency trigger condition association data, and transfer facility configuration data at each station in urban rail transit; The rail transit passenger flow data prediction module 103 is used to combine the collected real-time passenger flow data of historical urban rail transit and the associated data of emergency trigger conditions, design a transfer function between emergency events and passenger flow data, and calculate the impact value of emergency events on passenger flow data; based on the deep learning algorithm, construct a basic passenger flow prediction model, embed the emergency trigger model in the basic passenger flow prediction model to achieve adaptive adjustment of the prediction parameters in the basic passenger flow prediction model, and simultaneously add an event embedding layer containing emergency trigger conditions, emergency events and passenger flow data transfer function to the basic passenger flow prediction model to obtain the impact value of emergency events on passenger flow data and input it into the passenger flow prediction model to generate the passenger flow prediction model; and use the passenger flow prediction model to predict passenger flow data for each station; The rail transit transfer facility configuration acquisition module 104 is used to apply a multi-objective optimization algorithm to each station, combined with the predicted passenger flow data, to obtain the optimal transfer facility configuration strategy to minimize the sum of queue waiting costs and transfer facility operating resource costs over multiple time periods while satisfying transfer facility capacity constraints. The rail transit transfer facility configuration adjustment module 105 is used to adjust the transfer facility configuration for each station according to the configuration strategy of the optimal transfer facility, monitor the transfer efficiency of the station in real time, and continuously adjust the multi-objective optimization algorithm parameters or passenger flow prediction model parameters through the feedback mechanism until the transfer efficiency reaches the preset transfer efficiency.

[0051] In a specific embodiment, the system further includes: The rail transit transfer facility configuration verification module 106 is also used to construct a digital twin three-dimensional model of the station including the transfer facilities of the urban rail transit hub for each station; under the conditions of setting the predicted transfer flow demand and the configuration strategy of the optimal transfer facilities, use simulation technology to complete the transfer simulation and obtain the transfer efficiency of the transfer simulation process; after judging that the transfer efficiency of the transfer simulation process is less than the preset transfer efficiency, continuously adjust the multi-objective optimization algorithm parameters through the feedback mechanism until the transfer efficiency of the transfer simulation process is greater than the preset transfer efficiency, and use the adjusted multi-objective optimization algorithm parameters to re-obtain the configuration strategy of the optimal transfer facilities to replace the previously obtained optimal transfer facility configuration strategy adjustment.

[0052] In a specific embodiment, the rail transit data collection module 102 in the system is further configured to collect surrounding building data of each station in urban rail transit, and to extract GIS spatial features of each station using spatial analysis technology, including land use type and road network features; the land use type includes commercial land and residential land; the road network features include: high road network density, medium road network density, and low road network density; and is further configured to consider the service life of different types of transfer facilities at each station and obtain capacity depreciation coefficients of different types of transfer facilities within different service life periods. The rail transit transfer facility configuration acquisition module 104 is further used to pre-design a range of weights of different transfer facility capacity configuration proportions that match each GIS spatial feature combination according to different GIS spatial feature combinations, so as to satisfy that the spatial feature combination containing commercial land types has a higher weight of the capacity configuration proportion of elevator or escalator transfer facilities than the spatial feature combination containing residential land types, because the passenger flow in commercial areas is usually larger and the demand for vertical transportation is also higher; and the spatial feature combination containing high road network density has a higher weight of the capacity configuration proportion of transfer facilities than the spatial feature combination containing medium road network density. Compared with other transfer facility type combinations, the corresponding capacity weight of each facility type combination is higher. This is because the spatial characteristics of high road network density correspond to smaller transfer facility spacing requirements than those of medium road network density. Transfer facility combinations with smaller average distances are more likely to be selected by passengers, and the corresponding capacity weight of transfer facility combinations with smaller average distances is greater than the corresponding capacity weights of other transfer facility combinations. Similarly, the capacity weights of different transfer facility configurations that match spatial characteristics combinations with medium road network density are higher than those of spatial characteristics combinations with low road network density. The specific range of different transfer facility capacity configuration weights that match each GIS spatial characteristic combination is generated through deep learning model training using different GIS spatial characteristic combinations to determine the different transfer facility capacity weight ranges that meet the conditions. In the process of using a multi-objective optimization algorithm and combining predicted passenger flow data to obtain the optimal strategy for optimizing the configuration of transfer facilities, the weight range of the capacity configuration proportion of different transfer facilities in the configuration optimization strategy of transfer facilities is restricted to meet the matching weight range of the capacity configuration proportion of different transfer facilities; it is also used to introduce the capacity depreciation coefficient of the transfer facility into the said constraint condition and readjust the constraint condition.

[0053] The embodiment of the present application also discloses a computer-readable storage medium.

[0054] Specifically, the computer-readable storage medium stores a computer program that can be loaded by a processor and executed such as the above-mentioned method for optimizing the configuration of transfer facilities in an urban rail transit hub. The computer-readable storage medium includes, for example: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0055] The embodiment of the present application also discloses a computer device.

[0056] Specifically, the computer device includes a memory and a processor, and the memory stores a computer program that can be loaded by the processor and execute the above-mentioned method for optimizing the configuration of transfer facilities in urban rail transit hubs.

[0057] The above are all preferred embodiments of the present application and are not intended to limit the scope of protection of this application. Unless otherwise stated, any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features. In other words, unless otherwise stated, each feature is merely an example of a series of equivalent or similar features.

Claims

1. A method for optimizing the configuration of transfer facilities in an urban rail transit hub, characterized in that: include: Considering the influencing factors of emergencies in urban rail transit hub transfers, construct emergencies and their triggering conditions; Collect real-time passenger flow data, emergency trigger condition correlation data, and transfer facility configuration data for each station in urban rail transit; combine the collected historical real-time passenger flow data and emergency trigger condition correlation data in urban rail transit to design a transfer function between emergency events and passenger flow data, and calculate the impact of emergencies on passenger flow data; Based on a deep learning algorithm, a basic passenger flow prediction model is constructed. An emergency event trigger model is embedded in the basic passenger flow prediction model to achieve adaptive adjustment of the prediction parameters in the basic passenger flow prediction model. Simultaneously, an event embedding layer containing emergency event trigger conditions and a transfer function between emergency events and passenger flow data is added to the basic passenger flow prediction model to obtain the impact value of emergency events on passenger flow data and input it into the passenger flow prediction model to generate a passenger flow prediction model. The passenger flow prediction model is used to predict passenger flow data for each station. For each station, a multi-objective optimization algorithm is used, combined with predicted passenger flow data, to obtain the optimal transfer facility configuration strategy, minimizing the sum of queue waiting costs and transfer facility operating resource costs over multiple time periods while meeting transfer facility capacity constraints. For each station, the transfer facility configuration is adjusted according to the optimal transfer facility configuration strategy, and the transfer efficiency of the station is monitored in real time. The multi-objective optimization algorithm parameters or passenger flow prediction model parameters are continuously adjusted through the feedback mechanism until the transfer efficiency reaches the preset transfer efficiency.

2. The method for optimizing the configuration of transfer facilities in urban rail transit hubs according to claim 1, characterized in that: Also includes: Collect building data around each station in urban rail transit, and use spatial analysis technology to extract GIS spatial characteristics of each station, including land use type and road network characteristics; The land use types include commercial land and residential land; the road network characteristics include high road network density, medium road network density and low road network density; According to different GIS spatial feature combinations, different transfer facility capacity configuration ratio weight ranges matching each GIS spatial feature combination are pre-designed, so that the weight of the elevator or escalator transfer facility capacity configuration ratio in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing commercial land types is higher than that matching the spatial feature combination containing residential land types, and the transfer facility type combination with a smaller average distance in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing high road network density is higher than that matching the spatial feature combination containing medium road network density, and the transfer facility type combination with a smaller average distance in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing medium road network density is higher than that matching the spatial feature combination containing low road network density, and the transfer facility type combination with a smaller average distance in the different transfer facility capacity configuration ratio weights matching the spatial feature combination containing medium road network density is higher than that matching the spatial feature combination containing low road network density. In the process of using a multi-objective optimization algorithm and combining predicted passenger flow data to obtain the optimal strategy for optimizing the configuration of transfer facilities, the weight range of the capacity configuration proportion of different transfer facilities in the configuration optimization strategy of transfer facilities is restricted to meet the matching weight range of the capacity configuration proportion of different transfer facilities.

3. The method for optimizing the configuration of transfer facilities in urban rail transit hubs according to claim 2, characterized in that: The above method combines the real-time passenger flow data of historical urban rail transit and the associated data of emergency trigger conditions to design a transfer function between emergency events and passenger flow data, and calculates the impact value of emergency events on passenger flow data using the formula: ΔQ(t)=∑[f i (E i ,S i ,D i ,R i )*Q Base (t)] Where ΔQ(t) is the impact value of passenger flow data caused by the emergency at time t; f i (E i ,S i ,D i ,R i ) is the transfer function for the i-th sudden event, which depends on the event type E i , severity S i Duration D i And the impact range R i ; Among them, emergency type E i Including transfer facility failure at this station, facility failure at adjacent stations, suspension or addition of trunk stations, extreme weather or large-scale events and holidays; severity level S i Set to a value of 1-10 to indicate the urgency and severity of the emergency; duration D i It refers to the length of time from the beginning to the end of an emergency in minutes, hours or days; the impact range R i Refers to the scope of a single or several sites; Q Base (t) is the predicted passenger flow base value at time t when no emergency occurs.

4. The method for optimizing the configuration of transfer facilities in urban rail transit hubs according to claim 3, characterized in that: For each station, a multi-objective optimization algorithm is used, combined with the predicted passenger flow data, to obtain the configuration optimization strategy of the transfer facilities, so as to minimize the sum of the queue length and resource cost in multiple time periods, while satisfying the transfer facility capacity constraint and resource utilization constraint. The multi-objective optimization algorithm formula is as follows: The objective function is: Where W q (t) represents the waiting time of the queue in the t-th time period. The specific time period is obtained by artificial division based on a 24-hour day, and it is assumed that the transfer facility service process conforms to the queuing theory model and is determined according to the transfer flow and transfer facility configuration strategy; C(t) represents the transfer facility operation resource cost in the t-th time period, which is determined according to the transfer facility configuration strategy and the transfer facility operation unit cost; W queue and W cost Both represent weight coefficients; The constraints are: Transfer facility capacity ≥ Q predict (t) Where, the transfer facility capacity is obtained based on the configuration of the transfer facility and the time period t; Q predict (t) is the transfer traffic demand predicted in the tth time period; The decision variable is the configuration strategy P of the transfer facilities, which can be expressed as: P(K,M (K) ,X (K) ,V (K) ) K={k1,k2,……k n } In the formula, the configuration strategy P of transfer facilities depends on the type of transfer facilities K and the number of transfer facilities M under the condition of the triggering emergency event. (K) , the location X of the corresponding transfer facility under the condition of transfer facility selection K (K) , the operating speed V of the corresponding transfer facility under the condition of transfer facility selection K (K) ;k i represents the i-th type of transfer facility, where i ranges from 1 to n; Indicates the maximum number of the i-th transfer facilities at the current station; It shows the maximum safe operating rate of the i-th transfer facility at the current station.

5. The method for optimizing the configuration of transfer facilities in urban rail transit hubs according to claim 4, characterized in that: The configuration strategy range of the transfer facilities corresponding to the decision variable is dynamically adjusted according to the impact assessment level of the triggered emergency event; specifically, the impact assessment level of the triggered emergency event is determined based on the passenger flow quantity impact value range interval of the passenger flow data impact value caused by the calculated emergency event, and each passenger flow quantity impact value range interval corresponds to an impact assessment level of a type of emergency event; the impact assessment level includes major impact and non-major impact; In response to emergencies triggered by an impact assessment level of significant impact, the transfer facility types in the transfer facility configuration strategy are adjusted to add emergency transfer facility type options. The location, maximum number, and maximum safe operating rate of the transfer facilities corresponding to the newly added emergency transfer facility types are supplemented.

6. The method for optimizing the configuration of transfer facilities in urban rail transit hubs according to claim 4, characterized in that: Also includes: For each station, consider the service life of different types of transfer facilities at each station and obtain the capacity depreciation coefficient of different types of transfer facilities within different service life; Introducing the capacity depreciation coefficient of the transfer facilities into the constraints, the constraints are readjusted to: ∑ψ i * Under the condition of P, the capacity of the i-th type transfer facility in time period t ≥ Q predict (t) Where, ψ i Represents the capacity loss coefficient of the i-th type transfer facility.

7. The method for optimizing the configuration of transfer facilities in urban rail transit hubs according to claim 1, characterized in that: Also includes: For each station, a digital twin 3D model of the station including the transfer facilities of the urban rail transit hub is constructed; Under the conditions of setting the predicted transfer flow demand and the configuration strategy of the optimal transfer facilities, the transfer simulation is completed using simulation technology to obtain the transfer efficiency of the transfer simulation process; After determining that the transfer efficiency obtained in the transfer simulation process is lower than the preset transfer efficiency, the multi-objective optimization algorithm parameters are continuously adjusted through the feedback mechanism until the transfer efficiency obtained in the transfer simulation process is higher than the preset transfer efficiency, and the configuration strategy of the optimal transfer facility is re-obtained to replace the configuration strategy adjustment of the optimal transfer facility originally obtained.

8. An urban rail transit hub transfer facility optimization configuration system, characterized in that: include: Rail transit emergency event construction module, which is used to consider the influencing factors of emergency events at urban rail transit hub transfers and construct emergency events and their triggering conditions; Rail transit data collection module, used to collect real-time passenger flow data, emergency trigger condition correlation data, and transfer facility configuration data at each station in urban rail transit; The rail transit passenger flow data prediction module is used to combine the collected real-time passenger flow data of historical urban rail transit and the associated data of emergency trigger conditions, design the transfer function between emergency events and passenger flow data, and calculate the impact value of emergency events on passenger flow data; Based on a deep learning algorithm, a basic passenger flow prediction model is constructed. An emergency event trigger model is embedded in the basic passenger flow prediction model to achieve adaptive adjustment of the prediction parameters in the basic passenger flow prediction model. Simultaneously, an event embedding layer containing emergency event trigger conditions and a transfer function between emergency events and passenger flow data is added to the basic passenger flow prediction model to obtain the impact value of emergency events on passenger flow data and input it into the passenger flow prediction model to generate a passenger flow prediction model. The passenger flow prediction model is used to predict passenger flow data for each station. The rail transit transfer facility configuration acquisition module is used to apply a multi-objective optimization algorithm to each station, combined with predicted passenger flow data, to obtain the optimal transfer facility configuration strategy, thereby minimizing the sum of queue waiting costs and transfer facility operating resource costs over multiple time periods while meeting transfer facility capacity constraints. The rail transit transfer facility configuration adjustment module is used to adjust the transfer facility configuration for each station according to the optimal transfer facility configuration strategy, monitor the transfer efficiency of the station in real time, and continuously adjust the multi-objective optimization algorithm parameters or passenger flow prediction model parameters through the feedback mechanism until the transfer efficiency reaches the preset transfer efficiency.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method according to any one of claims 1 to 7.

10. A computer device, characterized in that: The computer device includes a memory, a processor, and a program stored and executable on the memory, and when the program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images