Energy-saving method and device for optimizing train operation timetable in real time

Abstract

The invention provides an energy-saving method and device for optimizing a train operation timetable in real time, and the method comprises the steps: collecting the operation data of a train in real time, and predicting the passenger flow of a plurality of stations in the future. And estimating the current running state of the train based on the running data and the passenger flow. And by taking the current running state of the train as initial input and taking minimization of the total traction energy consumption of all trains in a future optimization window as a target, constructing an optimization model and solving the optimization model to obtain a timetable adjustment parameter for energy-saving optimization of each train. And sending the solved timetable adjustment parameters to a corresponding train, and performing real-time adjustment on the train operation timetable. The train operation schedule can be dynamically optimized according to the operation data collected in real time and the predicted passenger flow, traction force and braking force are distributed more reasonably in the operation process of the train, unnecessary acceleration and deceleration are reduced, and therefore the traction energy consumption of the train is remarkably reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of train control technology, and in particular to a method and device for real-time optimization of train operation timetable for energy saving. BACKGROUND

[0002] With the increasing of the operation mileage of urban rail transit, the energy consumption of rail transit is increasing day by day, and reducing the train traction energy consumption can save a lot of energy. The energy saving method based on train operation timetable only needs to fine-tune the train operation timetable, without modifying the Automatic Train Supervision (ATS) system or the Automatic Train Operation (ATO) system, which is relatively simple to implement and has low cost. However, it cannot be adjusted in real time and has a lag. SUMMARY

[0003] The present application provides a method and device for real-time optimization of train operation timetable for energy saving, to solve the defect that the energy saving method based on train operation timetable in the prior art cannot be adjusted in real time and has a lag. The technical solution of the present application is as follows: In a first aspect, the present application provides a method for real-time optimization of train operation timetable for energy saving, comprising: real-time collection of train operation data and prediction of passenger flow at future stations; estimation of the current running state of the train based on the operation data and the passenger flow; taking the current running state of the train as the initial input, taking the minimization of the total traction energy consumption of all trains within a future optimization window as the target, constructing an optimization model and solving it to obtain the timetable adjustment parameters for energy saving optimization of each train; sending the solved timetable adjustment parameters to the corresponding train to adjust the train operation timetable in real time.

[0004] Optionally, the timetable adjustment parameters for energy saving optimization include the stop time adjustment amount of the train at the station and the running time adjustment amount of the train in the running section.

[0005] Optionally, the optimization model contains at least one of the following constraint conditions: safety interval constraint; stop time constraint; section running time constraint; train dynamics constraint; The safety interval constraint is that the distance between two vehicles at any time is greater than the minimum safety distance; the stop time constraint is that the adjusted stop time is within the allowed running time range; the interval running time constraint is that the deviation of the adjusted timetable does not exceed a preset threshold; and the train dynamics constraint is that the acceleration and deceleration are within the allowed range.

[0006] Optionally, the objective function of the optimization model is: ; Wherein, The minimization function is represented by min, The traction energy consumption of train k in running interval s is represented by Ek,s,k, k represents the train index, and s represents the running interval index, The set of all trains considered within the optimization window is represented by K, The set of all running intervals considered within the optimization window is represented by S.

[0007] Optionally, the traction energy consumption of the train is determined according to the train mass, acceleration, slope resistance, curve resistance and air resistance; The running data includes real-time full load rate, and the train mass is determined according to the train empty mass and the real-time full load rate.

[0008] Optionally, the optimization window is the running interval of the future several stations from the current time.

[0009] In a second aspect, the present application also provides an energy-saving device for real-time optimization of train running timetable, comprising the following modules: A data acquisition module is configured to acquire real-time running data of the train and predict passenger flow of future several stations; A state estimation module is configured to estimate the current running state of the train based on the running data and the passenger flow; An optimization solving module is configured to take the current running state of the train as the initial input, take minimizing the total traction energy consumption of all trains within a future optimization window as the target, construct an optimization model and solve it to obtain the timetable adjustment parameters of each train for energy-saving optimization; A communication issuing module is configured to send the solved timetable adjustment parameters to the corresponding train to real-time adjust the train running timetable.

[0010] In a third aspect, the present application also provides an electronic device comprising a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor implements the energy-saving method for real-time optimization of train running timetable as described in the first aspect.

[0011] In a fourth aspect, the present application also provides a non-transitory computer-readable storage medium having stored thereon a computer program, which, when executed by a processor, implements the energy-saving method for real-time optimization of train operation timetable according to the first aspect.

[0012] In a fifth aspect, the present application also provides a computer program product comprising a computer program, which, when executed by a processor, implements the energy-saving method for real-time optimization of train operation timetable according to the first aspect.

[0013] Based on the above technical solutions, the present application has the following beneficial effects compared with the prior art: The energy-saving method and device for real-time optimization of train operation timetable provided by the present application can obtain the latest status of train operation and the dynamic changes of passenger flow by collecting real-time train operation data and predicting passenger flow at future stations. This enables subsequent optimization and adjustment to be based on the latest and most accurate information, rather than being based on fixed and possibly outdated timetables as in the prior art, thereby providing a data basis for real-time adjustment. Based on the real-time collected operation data and predicted passenger flow, the current running state of the train is estimated. This real-time estimation can accurately reflect the actual running situation of the train at the moment, including the position, speed, and passenger load of the train, thereby providing accurate initial conditions for subsequent optimization and adjustment and avoiding the problem of adjustment lag caused by the use of outdated or inaccurate state information. The current running state of the train is taken as the initial input, and an optimization model is constructed and solved with the goal of minimizing the total traction energy consumption of all trains within a future optimization window. This approach not only considers the current actual situation of the train, but also focuses on the overall energy consumption optimization within a future time period, enabling forward-looking adjustment of the train operation timetable in real time. Compared with the method of the prior art, which cannot be adjusted in real time, this method can dynamically optimize the train operation timetable according to real-time conditions, plan a more energy-saving operation scheme in advance, and effectively overcome the problem of lag. The adjustment parameters of the solved timetable are sent to the corresponding train, and the train operation timetable is adjusted in real time. This step directly applies the optimization results to train operation, realizing real-time adjustment in a true sense. Unlike the method of the prior art, which is based on fixed timetables and cannot be updated in real time, this method can modify the train operation timetable in a timely manner based on real-time collected data and optimization results, ensuring that the train always operates according to the optimal energy-saving scheme, thereby effectively solving the problem of non-real-time and lag in the prior art.

[0014] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application will be realized and attained by the structure particularly pointed out in the description and the drawings.

[0015] In order to make the above objectives, characteristics and advantages of the present application more obvious and easy to understand, the following preferred embodiments are specifically described below with reference to the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor based on these drawings.

[0017] Figure 1 is a flowchart of the energy-saving method for real-time optimization of train operation timetable provided by the present application.

[0018] Figure 2 is a structural schematic diagram of the energy-saving device for real-time optimization of train operation timetable provided by the present application.

[0019] Figure 3 is a structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0020] In order to make the objectives, technical solutions and advantages of the present application more clear, the technical solutions of the present application will be described clearly and completely below with reference to the drawings in the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0021] Currently, there are three kinds of energy-saving methods based on train operation timetable. One is the running chart translation method, which only changes the train departure station time, without changing the train stop time and interval running time. One is the stop time adjustment method, which increases the traction-brake process matching by adjusting the stop time. One is the comprehensive adjustment method, which simultaneously considers the adjustment of departure interval and stop time. The above three methods cannot achieve real-time adjustment and have hysteresis.

[0022] The present application provides an energy-saving method and device for real-time optimization of train operation timetable, which adjusts the train interval running time and stop time in real time according to the full load rate of the train, the interval between the previous train and other information, so as to achieve the purpose of energy saving. Referring to Figure 1 The method comprises the following steps: S110, real-time collection of train operation data and prediction of passenger flow at future stations.

[0023] The above operation data is directly obtained from the ATS system and the on-board sensors. The operation data includes the position, speed and full load rate of the train, etc. The position is obtained through GPS, trackside transponders or odometer. The speed is obtained in real time through the speed sensor on the train. The full load rate is estimated through the car weight sensor (such as pressure sensor), video intelligent analysis (estimate the number of people) in the car or car door grating, etc. The full load rate is the ratio of the current passenger capacity to the rated passenger capacity. All data is transmitted in real time to the optimization server of the ground control center through the train-ground wireless communication network (such as LTE-M, Wi-Fi).

[0024] A passenger flow prediction model is pre-trained, which predicts the number of passengers getting on and off at each station in the future optimization window according to historical and multi-source real-time data. The input of the passenger flow prediction model is a high-dimensional, multi-source feature vector. The feature vector includes the following types of features: time features, historical passenger flow features, real-time state features and external environment features.

[0025] The time features include absolute time and relative time. The absolute time refers to year, month, day, hour, minute and day of the week. The "year-month-day hour: minute" is converted into Unix timestamp and binned into "hour" and "minute" two features. The relative time refers to whether it is a weekday, whether it is a holiday, the time period in a day (such as morning peak and evening peak), and the season in a year. The periodic features (such as hour) are encoded using sine or cosine. There are 24 hours in a day, and the passenger flow and operation demand of the train running in different hours will show periodic changes. Using sine or cosine encoding can convert the periodic time feature of hours into numerical form, which is convenient for processing and analysis in the optimization algorithm. Specifically, the 24 hours of a day can be mapped to a sine or cosine function with a period of 24, for example, for the hth hour, the following formula can be used for encoding: the encoding formula of sine coding is sin(2πh / 24). In this way, each hour will correspond to a unique sine and cosine value, which can reflect the periodic relationship and relative position between hours, providing a mathematical basis for the optimization of train operation timetable.

[0026] The historical passenger flow features include the target station history and the related station history. The target station history refers to the number of passengers getting on and off at the station in the past N time intervals (such as the past 6 15 minutes). N is dynamically adjusted according to weekdays or weekends (such as N=6 on weekdays and N=4 on weekends). The related station history refers to the historical passenger flow data of the upstream stations (such as the first 2 stations in the direction of train operation) at the same period, and if the data is missing, it is filled forward. The boarding and alighting of the upstream stations directly affects the load of the downstream stations.

[0027] Real-time state features include current short-term passenger flow, platform crowding, and train fullness. Current short-term passenger flow refers to the real-time number of passengers entering the station (from the gate) in the recent time interval. Platform crowding refers to the current number of waiting passengers on the platform based on video analysis estimates. Train fullness refers to the fullness of the train just leaving the station, which reflects whether the demand is fully met.

[0028] External environment features include weather conditions and special event flags. Weather conditions are categorical variables such as sunny, rainy, snowy, temperature, humidity. Special event flags are binary variables indicating whether there are known large-scale activities (such as concerts, sports events) occurring.

[0029] The output of the passenger flow prediction model is a continuous real value representing the predicted number of passengers. For example, given an input feature vector, the model outputs "A station boarding passengers," "A station alighting passengers," "B station boarding passengers," and so on. The training process of the passenger flow prediction model is performed offline on historical data sets, aiming to find the optimal values of model parameters that minimize prediction error. The training process of the model is as follows: 1. Data preparation and preprocessing: All mentioned input features and corresponding true passenger flow labels (i.e., real boarding and alighting passenger numbers recorded historically) are collected from the database. Categorical features (such as weather, day of the week) are converted into numerical form, with weather using One-Hot encoding and day of the week using sine or cosine encoding to preserve periodicity. Numerical features are standardized or normalized to be within similar numerical ranges, accelerating model convergence and improving performance. Training samples are constructed in a sliding window manner in chronological order. For example, use data at times [t-6, t-5,..., t-1] as features to predict passenger flow at time t.

[0030] 2. Model selection: Recurrent neural networks such as LSTM or GRU. Handle time series data, can capture long-term temporal dependencies well. It inputs features at each time step in turn, updates internal state step by step, and finally makes predictions.

[0031] 3. Model learning: The preprocessed data set is divided into training set, validation set and test set. The training set is used to learn parameters, the validation set is used to adjust hyperparameters and prevent overfitting during training, and the test set is used to evaluate the final performance of the model. The core algorithm is to minimize the loss function (such as mean square error MSE). For multi-output tasks, weighted MSE can be used, for example: loss function L = w 1 MSE 上车 + w 2 MSE 下客 . w 1 is the weight coefficient of the boarding passenger prediction task.w 2 is the weight coefficient of the down passenger number prediction task. MSE 上车 is the mean square error between the up passenger number prediction value and the true value. MSE 下客 is the mean square error between the down passenger number prediction value and the true value.

[0032] For LSTM, use the backpropagation through time algorithm. Starting from the prediction error, calculate the gradient of the loss function with respect to each weight, then use the optimizer (such as Adam) to update all parameters in the direction of gradient descent. This process is repeated several times through the training data, and an early stopping mechanism is set (such as the validation set RMSE does not decrease for 5 consecutive rounds, then stop training).

[0033] 4. Model evaluation and saving: Monitor model performance using the validation set. When the performance no longer improves, stop training in advance to prevent overfitting. Finally, evaluate the generalization ability of the model using the test set, calculate indicators such as mean absolute error (MAE), root mean square error (RMSE). Save the trained model structure and parameters as a file (such as.pkl,.h5,.onnx) for loading when predicting.

[0034] The prediction process is carried out online in real time, using the trained passenger flow prediction model and new data to generate prediction values. Real-time listening to message queues or databases, grabbing all input feature data at the current time (time, weather, real-time gate number, recent train full load rate, etc.). Apply the same preprocessing procedure to the real-time data as in the training phase. Load the previously saved model file. The preprocessed real-time feature vector is input into the model. The model performs a forward propagation calculation: the LSTM model calculates the current input and internal state vector, updates the state and outputs the prediction value.

[0035] S120, based on the running data and the passenger flow, estimate the current running state of the train.

[0036] Based on the real-time collected running data and the prediction data of the passenger flow prediction model, estimate the current running state of the train, including but not limited to the train's traction energy consumption, estimated arrival time. The estimated arrival time refers to the timestamp of predicting the train's arrival at the next station platform. According to the real-time speed, position (line slope, curvature) and full load rate, estimate the train's traction energy consumption. According to the current position, speed and the conditions of the line ahead, predict the train's arrival time at the station without changing speed. The running state estimation is carried out through a prediction-update cycle. The system maintains a train kinematics model internally. This model predicts the current state based on the state at the last time. The train kinematics model is: Position predicted = Position + v 0 x Δt +0.5× a ×Δ t 2 v = v 0+ a ×Δ t Passenger loadpredicted = Passenger loadprevious in, Position predicted This indicates the predicted position of the train at the next moment. Position This indicates the train's current position. v 0 indicates that in the time interval Δ t The initial speed of the train at the start, which is the current speed of the train. Δ t The time interval used for prediction represents the length of time elapsed from the current moment to the next moment. a For the train at time interval Δ t Acceleration within. Passenger loadpredicted The predicted passenger capacity of the train at the next moment is represented by the time interval Δ. t The expected number of passengers on the train. Passenger loadprevious This represents the number of passengers the train is carrying at the current moment.

[0037] Train traction energy consumption is calculated based on traction force, speed, and distance. Traction force F can be calculated based on acceleration and resistance (related to speed and weight): F = m × a + R(v). m Train quality includes both the empty car quality and the passenger quality. R ( v The resistance acting on the train (including gradient resistance, air resistance, etc.) is the resistance of the train speed. v The function.

[0038] The traction energy consumption model is as follows: E predicted = E previous +(F×v×Δt) / efficiency. E predicted The predicted energy consumption of the train at the next moment is represented by the time interval Δ. t The energy that the train is expected to consume. E previous This represents the train's energy consumption at the current moment. "Efficiency" indicates the efficiency of the train's traction system.

[0039] The weight of the train at station S can be calculated accurately according to the real-time full load rate when leaving station S: train mass = empty mass + (full load rate x average weight of each passenger x train capacity). According to the output of the passenger flow prediction model "the number of passengers boarding at S+1 station" and "the number of passengers alighting at S+1 station", it can be predicted that the passenger load of the train before arriving at S+1 station will remain unchanged, and after arriving at S+1 station, it will become: train mass = initial passenger load - number of passengers alighting + number of passengers boarding. This predicted train mass will be used as input for the train kinematics model and the traction energy consumption model until the next real-time full load rate data is received at the station. S+1 = initial passenger load S - number of passengers alighting S+1 + number of passengers boarding S+1 . This predicted train mass will be used as input for the train kinematics model and the traction energy consumption model until the next real-time full load rate data is received at the station.

[0040] Based on the current estimated position, speed, and remaining distance to the next station, combined with the slope of the line, speed limit, and train performance model, an integral operation is performed to predict the estimated arrival time. A set of state estimation values is output at each processing period (e.g., every second): the predicted arrival time of the train and the traction energy consumption.

[0041] S130, taking the current running state of the train as the initial input, and aiming to minimize the total traction energy consumption of all trains within a future optimization window, an optimization model is constructed and solved to obtain the timetable adjustment parameters for each train for energy-saving optimization.

[0042] The core objective of the optimization model is to minimize the total traction energy consumption of all trains within a future optimization window. Let there be n trains within the optimization window. For the k th train ( k = 1, 2, , n ), its running trajectory involves multiple stations and sections. The total traction energy consumption E total can be represented as the sum of the traction energy consumption of each train in each running section, is the traction energy consumption of train k in running section s.

[0043] The objective function of the optimization model is: where represents the minimization function, represents the traction energy consumption of train k in running section s, k represents the train index, s represents the running section index, represents the set of all trains considered within the optimization window, represents the set of all running sections considered within the optimization window.

[0044] Traction energy consumption of trains in the operating section With train travel time t k,s The train mass of train k in operating section s M k,s Operating resistance F r,k,s and train acceleration a k,s This is related to factors such as... According to the principles of train dynamics, it can be approximated as: in, F t,k,s ( t )yes t Time of the first k The train's traction force F r,k,s ( t )yes t Time of the first k The running resistance experienced by the train (including gradient resistance, curve resistance, etc.). v k,s ( t )yes t Time of the first k The speed of the train.

[0045] To simplify calculations, empirical formulas are established based on train operating characteristics and historical data, relating traction energy consumption to parameters such as travel time and train mass. For example, assuming the train operates in three phases—uniform acceleration, uniform speed, and uniform deceleration—within a given section, and the running resistance is constant, then the traction energy consumption can be simplified as follows: in, k 1 and k 2 is a constant related to train performance and operating conditions.

[0046] The decision variables of the above optimization model, that is, the timetable adjustment parameters used for energy-saving optimization, may include the adjustment amount of train stopping time at stations and the adjustment amount of train running time in the operating section.

[0047] The constraints of the above optimization model include stop time constraints, running time constraints, departure interval constraints, and passenger flow constraints.

[0048] The stop time constraint is the adjustment amount Δ for the train's stop time at the station. t s,k,q ( q The station number (representing the station's serial number) needs to meet certain upper and lower limits to ensure passengers have sufficient time to board and alight, without disrupting the station's normal operation. That is: in, t s,k,q It is the original stopping time of train k at station q. t s,min and t s,max These are the minimum and maximum allowed stopping times, respectively.

[0049] The running time constraint is the adjustment amount Δ of the train's running time within the operating section. t r,k,s Train travel time must not exceed a reasonable range to ensure that trains arrive at each station according to the prescribed routes and times. That is: in, t r,k,s It is the original travel time of the k-th train in the s-th operating section. t r,min and t r,max These are the minimum and maximum allowed running times, respectively.

[0050] Departure interval constraints: The departure interval between adjacent trains must meet certain safety requirements to avoid rear-end collisions and conflicts. Let the first train... k Train and the k The departure times for train +1 are as follows: T k and T k+1 Then the departure interval Δ T k = T k+1 T k Should meet: Where, Δ T min and Δ T max These are the minimum and maximum allowed departure intervals, respectively.

[0051] Passenger flow constraints: When adjusting train dwell times and travel times, the impact of passenger flow on train operation must be considered. For example, at stations with high passenger flow, appropriately increasing dwell time can reduce passenger waiting time and improve service quality; however, at stations with low passenger flow, excessively increasing dwell time will increase train energy consumption and travel time. Therefore, adjustments can be made based on predicted passenger flow. Pq (No. q Establish corresponding constraints for passenger flow at the station, such as: Where, f(P) q ) is a function related to passenger flow, representing the minimum stop time adjustment determined based on passenger flow.

[0052] This optimization model is a multivariable, multi-constraint nonlinear programming problem, which can be solved using intelligent optimization algorithms (such as genetic algorithms and particle swarm optimization algorithms). Taking the genetic algorithm as an example, the solution steps are as follows: (1) Coding: The stop time adjustment and running time adjustment of each train are coded to form a chromosome. For example, for a train with n Trains, each train passing by m Each operating range and m With +1 station, the chromosome length is... n ×(2 m +1).

[0053] (2) Initial population generation: A certain number of chromosomes are randomly generated to form the initial population.

[0054] (3) Fitness function design: Take the reciprocal of the objective function as the fitness function, i.e., f=1 / E total A higher fitness value indicates better individual performance.

[0055] (4) Selection operation: Based on the fitness value, a portion of individuals from the current population are selected as parents for the next generation using methods such as roulette wheel selection.

[0056] (5) Crossover operation: Perform crossover operation on the selected parent individuals to exchange some of their genes and produce new individuals.

[0057] (6) Mutation operation: Mutation operation is performed on new individuals with a certain probability to change some of their gene values ​​and increase the diversity of the population.

[0058] (7) Termination condition judgment: When the preset number of iterations or the fitness value meets certain accuracy requirements, the iteration stops and the optimal solution is output, which is the timetable adjustment parameter for each train used for energy saving optimization.

[0059] The present application can find the optimal adjustment scheme of train station stopping time and running interval running time by constructing an optimization model with the goal of minimizing total traction energy consumption and comprehensively considering various constraints of train operation. Compared with the method in the prior art which cannot be adjusted in real time, this method can dynamically optimize the train operation timetable according to the real-time collected operation data and the predicted passenger flow, so that the train can more reasonably allocate traction and braking forces during operation, reduce unnecessary acceleration and deceleration, thereby significantly reducing the traction energy consumption of the train and achieving the goal of energy saving and emission reduction. In the process of optimizing the train operation timetable, the influence of passenger flow on train operation is fully considered. By adjusting the train stopping time at the station, the waiting time of passengers can be reduced, the efficiency of passengers getting on and off the train can be improved, especially in stations with large passenger flow, the passenger congestion situation can be effectively alleviated, and the travel experience and service quality of passengers can be improved. Reducing the traction energy consumption of the train not only can reduce energy consumption and energy procurement cost, but also can reduce the wear and tear of train equipment and the failure rate, prolong the service life of the equipment, and reduce the maintenance and replacement cost of the equipment. At the same time, improving the efficiency and flexibility of train operation can reduce the situation of train delay and delay, which is also helpful to reduce the economic loss caused by operation interruption, thereby reducing the overall operation cost of urban rail transit.

[0060] The energy-saving method based on train operation timetable in the prior art cannot be adjusted in real time and has a lag. However, this method can adjust the train operation timetable in a timely manner according to the actual situation by collecting data in real time, predicting passenger flow, estimating train running state, and solving the optimization model in real time. This makes the train operation better adapt to the changes of passenger flow and the changes of line conditions, improves the real-time performance and flexibility of train operation, and reduces the problems of train delay and low running efficiency caused by unreasonable timetable.

[0061] S140, sending the solved timetable adjustment parameters to the corresponding train to adjust the train operation timetable in real time.

[0062] After obtaining the timetable adjustment parameters, these parameters need to be integrated first. The parameters such as the stopping time adjustment amount and the running time adjustment amount corresponding to each train are classified and arranged according to the train number to form a complete data set. For example, for train k , the data set should include the stopping time adjustment amount Δ t s,k,q ( q for the station number) and the running time adjustment amount Δ t r,k,s ( s for the running interval number).

[0063] The location and status of the target train that needs to receive timetable adjustment parameters are determined using the ATS system or other positioning equipment. The ATS system can acquire real-time information such as the train's running position, speed, and direction, and associate this information with the train number. For example, when train adjustments are needed... k When the timetable is accessed, the ATS system can quickly locate the train. k The current station or operating section is used to send timetable adjustment parameters to the train.

[0064] The timetable adjustment parameters are stored in the train's onboard control system. This system includes the Automatic Train Control (ATC) system and train operation monitoring and recording devices. These systems integrate and update these adjustment parameters with the existing timetable. For example, the ATC system recalculates the train's arrival and departure times at each station and its speed curves in each operating section based on the new stop time and travel time adjustments. The updated operating plan is then sent to the train's traction, braking, and other subsystems so that the train can operate according to the new timetable.

[0065] The train's onboard control system adjusts parameters based on the updated timetable to replan train operations. Regarding adjustments to stop times, if the stop time increases, the train will appropriately extend its stop time at stations to ensure passengers have sufficient time to board and alight; if the stop time decreases, the train will close its doors earlier and depart as quickly as possible. Regarding adjustments to travel time, if the travel time increases, the train will appropriately reduce its speed within the operating section to reduce traction and energy consumption; if the travel time decreases, the train will appropriately increase its speed, but must ensure that it does not exceed the line's maximum permissible speed and the train's performance limitations.

[0066] The energy-saving method for real-time optimization of train operation timetable provided by the present application can obtain the latest status of train operation and the dynamic changes of passenger flow in time by collecting real-time train operation data and predicting passenger flow in future stations. This enables subsequent optimization and adjustment to be based on the latest and most accurate information, rather than being based on fixed and possibly outdated timetables as in the prior art, thereby providing a data basis for real-time adjustment. Based on the real-time collected operation data and predicted passenger flow, the current operation state of the train is estimated. This real-time estimation can accurately reflect the actual operation status of the train at the moment, including the position, speed, passenger load and other key information, thereby providing accurate initial conditions for subsequent optimization and adjustment and avoiding the problem of adjustment lag caused by the use of outdated or inaccurate state information. The current operation state of the train is taken as the initial input, and an optimization model is constructed and solved with the goal of minimizing the total traction energy consumption of all trains within a future optimization window. This method not only considers the current actual situation of the train, but also focuses on the overall energy consumption optimization within a future time period, and can make forward-looking adjustments to the train operation timetable in real time. Compared with the method in the prior art that cannot be adjusted in real time, this method can dynamically optimize the train operation timetable according to real-time conditions, plan a more energy-saving operation scheme in advance, and effectively overcome the problem of lag. The adjustment parameters of the solved timetable are sent to the corresponding train, and the train operation timetable is adjusted in real time. This step directly applies the optimization results to train operation, realizing real-time adjustment in a true sense. Unlike the method in the prior art that is based on fixed timetables and cannot be updated in real time, this method can modify the train operation timetable in time according to real-time collected data and optimization results, ensure that the train always operates according to the optimal energy-saving scheme, and effectively solve the problem of non-real-time and lag in the prior art.

[0067] In an embodiment, the optimization model comprises at least one of the following constraint conditions: a safety interval constraint; a stop time constraint; an interval operation time constraint; a train dynamics constraint; The safety interval constraint is that the distance between two trains at any time is greater than the minimum safety distance. The stopping time constraint is that the adjusted stopping time is within the allowable running time range, which is determined according to factors such as passenger flow at the station, platform facility conditions, and average time for passengers to get on and off the train. The interval running time constraint is that the deviation of the adjusted timetable does not exceed a preset threshold. The preset threshold is determined according to factors such as operation requirements of the line, travel habits of passengers, and connection with other modes of transport. The train dynamics constraint is that the acceleration and deceleration are within the allowable range. The allowable acceleration and deceleration range is determined according to factors such as performance of the train, comfort requirements of passengers, and line conditions.

[0068] By setting the safety interval constraint, the train braking performance, line conditions, and possible emergencies are fully considered. In the process of timetable adjustment, strictly following this constraint can ensure that the front and rear trains maintain sufficient safety distance at any time, effectively avoiding serious safety accidents such as rear-end collisions. For example, when adverse weather (such as heavy rain and snow) causes the train braking distance to increase, the safety interval constraint can ensure that the train has enough braking space to protect the safety of passengers and the train. On complex lines with multiple trains running simultaneously, the passenger flow density, running speed, and line conditions differ greatly in different sections. The safety interval constraint can dynamically adapt to these changes to ensure the safety interval between trains in various operating scenarios. For example, during peak hours, the train running density is high, and the safety interval constraint can prevent safety hazards caused by too small train intervals; during off-peak hours, the train interval can also be reasonably controlled to improve the utilization efficiency of the line.

[0069] By setting the stopping time constraint, it can be ensured that the stopping time of the train at the station can meet the needs of passengers to get on and off the train safely and orderly, and will not affect the overall running efficiency of the train due to too long stopping time. For example, in large transfer stations, the passenger flow is large, and the stopping time needs to be appropriately extended to ensure that passengers have enough time to complete the transfer; while in small stations with small passenger flow, the stopping time can be shortened to improve the running speed of the train.

[0070] By setting the section running time constraint, the train can run according to the predetermined timetable, reduce the train's late and early situation, and improve the punctuality of the rail transit system. For example, for the line connected with the transportation hub such as airport and train station, the punctual running of the train is particularly important, and the section running time constraint can ensure that passengers can arrive at the destination on time and facilitate the transfer to other transportation tools. On the line with multiple trains running, the section running time constraint helps to maintain the stability of the entire operation order. If the section running time deviation of a train is too large, it may affect the running of the subsequent train, leading to the chaos of the entire line operation. By strictly controlling the deviation of the section running time, the running interval between trains can be ensured to be uniform, and the passing capacity and operation efficiency of the line can be improved.

[0071] By setting the train dynamics constraint, the train can maintain stability during acceleration and deceleration, reduce the sway and impact of the train, and improve the riding comfort of passengers. For example, excessive acceleration and deceleration will make passengers feel uncomfortable, and even may cause safety problems; while too small acceleration and deceleration will affect the running efficiency and energy saving effect of the train. The train dynamics constraint is closely related to energy saving optimization. Under the premise of meeting the train running stability and equipment protection, reasonable control of the acceleration and deceleration of the train can achieve energy saving operation. For example, using the slow start and slow stop running mode can avoid the sudden acceleration and deceleration of the train, reduce the traction energy consumption and brake energy loss, and improve the energy utilization efficiency of the train.

[0072] In an embodiment, the traction energy consumption is related to the train mass (including passenger mass), acceleration, speed, and line resistance (slope, curvature, and air resistance). The real-time load factor directly affects the train mass, which will affect the size of the traction force and braking force, thereby affecting the energy consumption. According to the train position and the load factor, the running time and the stop time are dynamically adjusted, so that the train runs on the energy consumption optimal curve under the constraint condition. The train's traction energy consumption is determined according to the train mass, acceleration, slope resistance, curve resistance, and air resistance. The running data includes the real-time load factor, and the train mass is determined according to the train empty mass and the real-time load factor.

[0073] The train mass is composed of two parts: empty car mass (i.e. empty mass) and passenger mass. Among them, the passenger mass is calculated by the real-time load factor. Let the maximum passenger capacity of the train be M max , the average body weight be m avg , and the real-time load factor be λ , then the passenger mass M passenger = λ × M max × m avg. So the train mass m = m empty + M passenger = m empty + λ × M max × m avg , m empty is the empty mass of the train.

[0074] Grade resistance R grade is related to the grade of the track. Different grades result in different resistances. Generally, grade resistance is positive when going uphill, hindering the train's progress; it is negative when going downhill. Its specific value can be calculated through the grade data of the track and relevant physical formulas. R curve Curve resistance R air is related to the curvature of the track. When the train passes through a curved track, it will generate additional resistance due to centrifugal force. The greater the curvature, the greater the curve resistance. Its calculation needs to consider factors such as the wheel-rail relationship of the train and the curve radius.

[0075] According to Newton's second law F = ma (where F is the traction force, m is the train mass, a is the acceleration), combined with the train mass m calculated earlier and the known acceleration a , the required traction force of the train can be initially calculated. At the same time, it also needs to consider the grade resistance R grade , curve resistance R curve , and air resistance R air that the train is subjected to. The actual traction force F needs to overcome these resistances and provide the force required for the train to accelerate, that is, .

[0076] The instantaneous traction power P is equal to the product of the traction force F and the train's running speed v , that isP = F × v .

[0077] The traction energy consumption is the integral of power over time. In a specific operating interval or time period [t1, t2], the traction energy consumption E in that interval or time period can be obtained by integrating the instantaneous traction power P(t). The traction energy consumption model is: In practical applications, approximate calculation can be performed by discretization method, dividing the operating time into multiple small time periods, calculating the energy consumption of each time period and summing them up.

[0078] The present application determines the train mass according to the real-time full load rate, can accurately calculate the traction energy consumption of the train, so that the train operation timetable can be optimized and adjusted in real time, so that the train can run on the energy consumption optimal curve under the condition of meeting various constraint conditions, thereby realizing precise energy saving effect. For example, according to the full load rate of the train in different time periods, the running time and stopping time of the train are dynamically adjusted to avoid unnecessary traction energy consumption waste. Through traction energy consumption calculation, the energy consumption of the train in different operating states can be determined. By optimizing the train operation timetable in real time, the train can run in the most energy-saving way under different line conditions and load conditions, improving the overall energy utilization efficiency. For example, during the flat peak period, the train full load rate is low, the running speed and stopping time can be adjusted appropriately to reduce the traction energy consumption; during the peak period, the train running interval and stopping time are reasonably allocated according to the full load rate to avoid energy loss caused by frequent acceleration and deceleration of the train, thereby improving the energy utilization efficiency.

[0079] The train mass and traction energy consumption are determined based on real-time full load rate, so that the optimization of train operation timetable can be dynamically adjusted according to the actual passenger flow. This flexibility enables the train to better adapt to passenger flow changes in different time periods and different sections, improving the operating efficiency and service quality of the rail transit system. For example, in the case of sudden large passenger flow, the running parameters of the train can be adjusted in time to increase the carrying capacity of the train while ensuring energy-saving operation; when the passenger flow is small, the running energy consumption of the train is reduced to realize reasonable allocation of resources.

[0080] In an embodiment, the optimization window is the operating interval of future stations from the current time. The process of solving the timetable adjustment parameters according to the optimization window is as follows: (1) Taking the current time as the starting point, determine the future stations (for example, the next 3 stations) as the optimization window. In this window, through the sensors on the train and the monitoring equipment of the rail transit system, collect all the real-time position, speed and full load rate of the train in the window.

[0081] For example, assume the current time is T0, and the optimization window contains stations A, B, and C, then the real-time position, speed, and full load rate of the train at T0 and after the train travels to A, B, and C need to be collected. The full load rate can be obtained by passenger flow counting equipment on the train, the real-time position can be determined by means of a GPS positioning system or track circuit positioning technology, and the speed is measured by a speed sensor.

[0082] (2) State estimation: According to the collected real-time data, the running state of the current train in the optimization window is estimated. This includes but is not limited to the current traction energy consumption and the predicted arrival time of the train, which can be estimated by the previous traction energy consumption model combined with real-time position, speed, and full load rate data. The predicted arrival time is predicted according to the current running speed of the train and the distance to the next station, combined with the line conditions (such as slope, curvature, etc.). For example, given that the train is currently 500 meters away from the next station, the current speed is 30 meters per second, and the line slope is 2%, the traction energy consumption and the predicted arrival time of the train at the next station can be estimated by the train dynamics model and the traction energy consumption model.

[0083] (3) Optimization model solving, which includes: Speed calculation: According to the safety constraint condition, the safe speed of the train in the optimization window is calculated. The safety interval constraint requires that the distance between any two trains at any time be greater than the minimum safety distance, and the position relationship between the two trains can be determined by real-time position information, and then the maximum allowable speed of the train that meets the safety distance requirement can be calculated. For example, if the distance between the rear train and the front train is 800 meters, the minimum safety distance is 300 meters, and the speed of the front train is 25 meters per second, the safe speed range of the rear train can be calculated according to the relative motion relationship.

[0084] Stop time adjustment range determination: For each train, the stop time adjustment range is determined according to its full load rate and current early or late time. Trains with high full load rates may need to shorten the stop time appropriately to improve efficiency, while trains with low full load rates can extend the stop time appropriately. At the same time, considering the current early or late situation of the train, if the train is late, it may need to reduce the stop time to recover time; if the train is early, it can appropriately increase the stop time. For example, a train with a full load rate of 90% and a late time of 2 minutes, its stop time adjustment range may be limited to 10-30 seconds shorter than the original plan; while a train with a full load rate of 50% and an early time of 1 minute, the stop time can be appropriately extended by 10-20 seconds.

[0085] Interval running time adjustment amount and energy-saving speed curve, traction energy calculation: for each train's interval running time adjustment amount, calculate the corresponding energy-saving speed curve and traction energy. According to the train mass (determined by the full load rate), line conditions (slope, curve resistance, etc.) and adjusted interval running time, the train dynamics model and energy consumption calculation model are used to calculate the energy-saving speed curve and traction energy of the train under the interval running time adjustment amount. For example, if the train's running time in a certain interval is extended by 10 seconds, the corresponding energy-saving speed curve can be obtained through model calculation, so that the train's energy consumption is minimized under this running time, and the traction energy at this time is calculated.

[0086] Combination optimization: considering all trains in the optimization window, select a set of stop time adjustment amount and interval running time adjustment amount to minimize the total traction energy of all trains in the adjustment window, while meeting all constraint conditions such as safety interval constraint, stop time constraint, interval running time constraint and train dynamics constraint. This can be achieved through mathematical optimization algorithms such as linear programming, nonlinear programming or heuristic algorithms, etc. For example, using genetic algorithm, through continuous iteration and optimization, find the adjustment scheme that meets all constraint conditions and has the minimum total traction energy.

[0087] (4) Instruction issuing: send the stop time adjustment amount and interval running time adjustment amount obtained by solving the optimization problem to the train. After receiving these instructions, the train runs according to the new timetable, realizing real-time adjustment of the train running timetable. For example, send the adjustment instructions to the on-board controller on the train through the wireless communication system, and the on-board controller controls the train's running speed and stop time according to the instructions.

[0088] Suppose two trains (A and B) are running on the line, A has a full load rate of 90%, B has a full load rate of 50%, A is about to enter the station, and B is behind A. Before adjustment: A and B stop according to the timetable; After adjustment: A (high full load rate) uses sliding mode when entering the station, reduces speed, extends interval running time, shortens stop time, and makes up for the loss of extended interval running time; B (low full load rate) can extend the stop time appropriately and maintain a safe distance from A.

[0089] The present application takes the future several stations from the current time as the optimization window, can adjust the train interval running time and stop time according to the real-time train operation data and full load rate, so that the train runs on the energy optimal curve under the condition of meeting various constraints. Compared with the traditional energy-saving method based on train operation timetable, this real-time optimization method can more accurately adapt to the actual running condition of the train, avoid the problem of poor energy-saving effect caused by data lag, and thus realize more effective energy-saving. For example, in the case of sudden passenger flow changes, the train operation parameters can be adjusted in time to reduce unnecessary traction energy consumption. By optimizing the train operation timetable in real time, the train running interval and stop time can be reasonably arranged to reduce the waiting time and empty running time of the train and improve the overall operation efficiency of the rail transit system. Within the optimization window, according to the real-time position and speed information of the train, the arrival and departure time of the train can be more accurately predicted to avoid conflicts and delays between trains, so that the train can run in a more reasonable order and time. For example, the situation of long waiting time of the following train caused by the late arrival of the previous train is avoided, and the line capacity is improved. This optimization method can better adapt to the passenger flow changes and line conditions in different time periods and different sections. In the peak period, the train running frequency can be increased and the stop time can be shortened according to the real-time full load rate to meet the travel demand of a large number of passengers; in the flat peak period, the train running interval and stop time can be appropriately extended to reduce energy consumption. At the same time, for different line conditions sections, such as sections with large slope or many curves, the train running speed and time can be adjusted according to the actual situation to ensure the safe and efficient operation of the train.

[0090] Optimizing the train operation timetable in real time can reduce the train delay and improve the punctuality rate of the train to provide more reliable travel services for passengers. Passengers can more accurately estimate their travel time and arrange their schedule reasonably. In addition, by optimizing the stop time, the train can be avoided to stay at the station for a long time, the waiting time of passengers is reduced, and the travel experience of passengers is improved. For example, when waiting for the train, passengers can clearly know the arrival time and stop time of the train, reducing uncertainty and anxiety. Since the energy-saving operation of the train is realized, the traction energy consumption is reduced, thereby reducing the energy procurement cost. At the same time, by improving the operation efficiency and reducing the equipment wear and tear, the maintenance and replacement cost of the equipment is reduced. For example, the equipment wear and tear caused by frequent acceleration and deceleration and long running of the train is reduced, the service life of the equipment is prolonged, and the operation cost is reduced. In addition, this optimization method does not need to make large-scale modification to the existing system, only needs to upgrade the software and optimize the algorithm based on the existing system, thereby reducing the cost of system upgrade.

[0091] The energy-saving device for real-time optimization of train timetables provided by the present invention will be described below. The energy-saving device for real-time optimization of train timetables described below can be referred to in correspondence with the energy-saving method for real-time optimization of train timetables described above.

[0092] The energy-saving device for real-time optimization of train timetables provided by this invention refers to... Figure 2 As shown, it includes: The data acquisition module 210 is used to collect train operation data in real time and predict passenger flow at several stations in the future. The status estimation module 220 is used to estimate the current operating status of the train based on the operating data and the passenger flow. The optimization solution module 230 is used to construct and solve an optimization model with the current operating state of the train as the initial input and the goal of minimizing the total traction energy consumption of all trains in a future optimization window, so as to obtain the timetable adjustment parameters for energy-saving optimization of each train. The communication sending module 240 is used to send the timetable adjustment parameters obtained by the solution to the corresponding trains to adjust the train timetable in real time.

[0093] Figure 3 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 3 As shown, the electronic device may include a processor 310, a communications interface 320, a memory 330, and a communication bus 340. The processor 310, communications interface 320, and memory 330 communicate with each other via the communication bus 340. The processor 310 can call logical instructions from the memory 330 to execute energy-saving methods for real-time optimization of train timetables.

[0094] In addition, the logic instructions in the memory 330 described above can be realized in the form of a software function unit and sold or used as an independent product, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that contribute to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium, and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0095] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program can be executed by a processor to enable a computer to execute the energy-saving method for real-time optimization of train operation timetable provided by the above-mentioned methods.

[0096] In another aspect, the present application also provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the energy-saving method for real-time optimization of train operation timetable provided by the above-mentioned methods.

[0097] The device embodiments described above are only schematic, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., they can be located in one place, or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment. Those skilled in the art can understand and implement without creative labor.

[0098] From the above description of the embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software plus necessary general hardware platforms, and of course can also be realized by hardware. Based on such understanding, the above technical solutions essentially or the parts that contribute to the prior art can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0099] Finally, it should be noted that the above examples are only used to illustrate the technical solutions of the present application, and are not intended to limit the same; although the present application has been described in detail with reference to the foregoing examples, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacements for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for optimizing energy saving of a train operation timetable in real time, characterized in that, The method comprises: collecting train operation data in real time and predicting passenger flow at future stations; estimating the current operation state of the train based on the operation data and the passenger flow; taking the current operation state of the train as initial input, constructing an optimization model and solving it to obtain time table adjustment parameters for energy-saving optimization of each train, with the goal of minimizing the total traction energy consumption of all trains within a future optimization window; sending the solved time table adjustment parameters to the corresponding trains to adjust the train operation time table in real time.

2. The energy saving method for real-time optimization of train schedules according to claim 1, characterized in that, The time table adjustment parameters for energy-saving optimization include train stop time adjustment amount at stations and train running time adjustment amount in running sections.

3. The energy saving method for optimizing train schedules in real time according to claim 1, characterized in that, The optimization model includes at least one of the following constraint conditions: safe interval constraint; stop time constraint; section running time constraint; train dynamics constraint; The safe interval constraint is that the distance between two trains at any time is greater than the minimum safe distance; the stop time constraint is that the adjusted stop time is within the allowed running time range; the section running time constraint is that the deviation of the adjusted time table does not exceed the preset threshold; and the train dynamics constraint is that the acceleration and deceleration are within the allowed range.

4. The energy saving method for optimizing train schedules in real time according to claim 1, characterized in that, The objective function of the optimization model is: ； wherein, denotes a minimization function, denotes the traction energy consumption of train k on running section s, k denotes the train index, s denotes the running section index, denotes the set of all trains considered within the optimization window, denotes the set of all running sections considered within the optimization window.

5. The energy saving method for real-time optimization of train schedules of claim 4, wherein, The train traction energy consumption is determined according to train mass, acceleration, slope resistance, curve resistance and air resistance; The operation data includes real-time full load rate, and the train mass is determined according to the train empty mass and the real-time full load rate.

6. The energy saving method for optimizing train schedules in real time according to claim 1, wherein, The optimization window is the running section of future stations from the current time.

7. An energy saving device for optimizing train operation schedule in real time, characterized in that, The method comprises: a data collection module for collecting train operation data in real time and predicting passenger flow at future stations; a state estimation module for estimating the current operation state of the train based on the operation data and the passenger flow; an optimization solving module for taking the current operation state of the train as initial input, constructing an optimization model and solving it to obtain time table adjustment parameters for energy-saving optimization of each train, with the goal of minimizing the total traction energy consumption of all trains within a future optimization window; a communication issuing module for sending the solved time table adjustment parameters to the corresponding trains to adjust the train operation time table in real time.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to realize the energy-saving method for real-time optimization of train operation time table according to any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the energy-saving method for real-time optimization of train operation time table according to any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to realize the energy-saving method for real-time optimization of train operation time table according to any one of claims 1 to 6.

Images