CN111401643A

CN111401643A - Urban rail transit passenger flow loop self-adaptive intelligent train scheduling method

Info

Publication number: CN111401643A
Application number: CN202010197173.2A
Authority: CN
Inventors: 颜红慧; 周庭梁; 钱江; 陈卫华
Original assignee: Casco Signal Ltd
Current assignee: Casco Signal Ltd
Priority date: 2020-03-19
Filing date: 2020-03-19
Publication date: 2020-07-10
Anticipated expiration: 2040-03-19
Also published as: CN111401643B

Abstract

The invention relates to an urban rail transit passenger flow loop self-adaptive intelligent train dispatching method, which comprises the following steps: step 1), predicting the off-train passenger capacity; step 2), acquiring the outbound capacity of passengers at a station; and 3) carrying out train operation capacity adjustment according to the information in the step 1) and the step 2. Compared with the prior art, the method has the advantages of complete support information, high informatization and intellectualization levels, high dispatching and commanding efficiency and the like.

Description

Urban rail transit passenger flow loop self-adaptive intelligent train scheduling method

Technical Field

The invention relates to an intelligent train dispatching method, in particular to an intelligent train dispatching method for self-adaption of a passenger flow loop of urban rail transit.

Background

Urban rail transit, as a high-speed and large-capacity public transport, is becoming increasingly important in mass travel, and especially when large-scale activities of several consecutive days exist, it is responsible for indelible.

When a large passenger flow occurs on a line due to equipment failure and the like, at present, a dispatcher can only check CCTV pictures by looking up the CCTV pictures, the number of the CCTV pictures which can be checked by a control center at the same time is limited, the dispatcher cannot conveniently master passenger flow information in time, and the dispatcher selects an operation control strategy according to ordinary experience.

With the continuous enrichment of passenger flow detection means and the rapid development of AI and big data technologies, how to effectively apply the technologies to the accurate acquisition of passenger flow information, so that the technology is a technical problem to be solved by adopting a matching strategy to process emergency situations and guide passenger flow in due time.

Disclosure of Invention

The invention aims to overcome the defects in the prior art and provide an intelligent train scheduling method for the urban rail transit passenger flow loop self-adaptation, which has the advantages of complete support information, high informatization and intelligentization levels, high scheduling command efficiency and the like and can assist a dispatcher in making a train operation scheduling decision when holding continuous large-scale activities for days.

The purpose of the invention can be realized by the following technical scheme:

an urban rail transit passenger flow loop self-adaptive intelligent train dispatching method comprises the following steps:

step 1), predicting the off-train passenger capacity;

step 2), acquiring the outbound capacity of passengers at a station;

and 3) carrying out train operation capacity adjustment according to the information in the step 1) and the step 2.

Preferably, the under-train passenger capacity prediction specifically includes:

step 1.1), carrying out big data analysis based on ticket service clearing data, train weighing and other data, and establishing an OD model of train departure load, passenger number of getting-on and getting-off and passenger flow of each station aiming at a current schedule;

step 1.2), if a certain station C_iThe exit is connected with a large-scale event place, large-scale events are held within a plurality of days continuously, the safety check of personnel going out of the station and entering the place is required, partial exits are closed according to the place reception capacity, and at the moment, the passenger arrival rate is controlled within the range of the passenger outbound capacity of the station, and the safety check of the personnel going out of the station and entering the place is requiredAnd predicting the passenger capacity under the train.

Preferably, the predicting the passenger capacity under the train specifically comprises:

step 1.2.1) passenger arrival rate prediction preconditions: station C in the day_iOther stations have no similar large-scale activities;

step 1.2.2) under normal conditions, at C_iWithin the range of n upstream stations, if the arrival intervals of the trains and the front vehicle on the same platform accord with the planned interval and the arrival intervals of the trains and the front vehicle on the same platform on the upstream of the current direction also accord with the planned interval, the departure load of the trains and the arrival intervals of the trains on the C station can be determined according to the departure load of the trains_i-n+1To C_i-1The number of passengers getting on and off each station between stations and the passenger flow model are compared with the train at C_iIf the number of passengers is predicted, if the station C_iNot a terminal station, and an uplink train and a downlink train are counted;

step 1.2.3) if C_jThe station has shielding door fault, j belongs to [ i-n +1, i-1]Train at C_jThe station adopts the over-station adjustment strategy, then C_jThe number of passengers getting on and off the station is 0, and the passengers on the vehicle can select to be in the position C_j-1Station and C_j+1Getting off the bus, if j is i-1, the bus needs to be checked at the bus station C_iMake an announcement that the number of newly added visitors is likely to be in C_iStanding out;

step 1.2.4) if the fault treatment results in that the threshold value F (actual departure interval-planned departure interval) of the train A and the previous train are on the same platform, the part of the train on the upstream of the train A is buckled and stopped at a late point, the part of the train on the upstream of the train A and the previous train is inserted for reducing the large interval and is adjacent to the passenger train, the passenger carrying conditions of the adjacent passenger train, the train A and the train at the late point on the upstream are different from those of the normal alignment point, and the part of the train is required to be at the station C_iThe number of passengers getting off the elevator is specifically predicted.

Preferably, the statistics of the number of people in the step 1.2.2) are specifically as follows:

step 1.2.2.1) R_j,j+1For train A at station C_jTo C_j+1Load of interval, j belongs to [ i-n, i-1 ]]The passenger flow is obtained by a multi-element passenger flow detection mode such as train weighing;

step 1.2.2.2) X_kFor train A at station C_kThe number of passengers, k, belongs to [ j +1, i-1 ]]The model value or the number of passengers getting off the train at the station can be referred;

step 1.2.2.3) S_kFor train A at station C_kThe number of passengers getting on, k belongs to [ j +1, i-1 ]]The number of passengers getting on the train can be predicted according to the model value or the number of the recent passengers getting on the train stopping at the station; train A is at C_kThe number of the passengers on the station is the smaller value of the predicted number of the passengers on the train A and the number of the passengers on the train A;

step 1.2.2.4) according to train on C_iPredicted arrival load, model arrival load, and model number of passengers in train_iThe number of passengers getting off, the train is at C_kK is [ i-n +1, i ]]The system can detect passenger flow in a multivariate passenger flow detection mode such as train weighing;

step 1.2.2.5) station C_i-nTo C_i-1Each train leaving the station of updating C_iThe number of passengers is predicted.

Preferably, the specific prediction in step 1.2.4) is as follows:

let C be_mFor the station when the train returns to operation, the next train is the starting station for inserting operation, and m belongs to [ i-1, 1]；

Step 1.2.4.1) estimating the number of regular passengers who do not go to a large-scale activity place in the departure load when the train leaves the station and the train A at the upstream late point and the number of passengers getting off the station of the regular passenger flow at the downstream station of the regular passenger flow;

step 1.2.4.2) prediction of C_iTrain in the range of upstream n stations is at C_iThe number of passengers getting off, if the station C_iNot the terminal station, the up and down trains are counted.

Preferably, the step 1.2.4.1) is specifically:

a) estimating train A and upstream late train C_mThe conventional number of persons who do not go to the large-scale event place in the load of the upstream station leaving the station and the number of persons who get off the large-scale event place in the part of conventional passenger flow at the respective downstream stations are as follows:

without going to large-scale activity places in current time and directionConventional passenger flow OD model: at station C_xAmong passengers getting on will have B_x:x+y% of passengers in C_x+yStanding passenger, y>＝1；

According to the OD model, the train is at the station C_mModel number of guests on each station at upstream calculation station C_m-1Model off-station load D_m-1Number of visitors under model of downstream station in current direction X_Dm-1:m-1+p，p>＝1；

Suppose that the current train is at station C_mThe number of guests at each upstream station includes the number of specific guests going to the large event and the number of model guests not going to the large event, and C_m-1Model off-station load D_m-1Number of visitors under model of downstream station in current direction X_Dm-1:m-1+pAs the current train C_m-1Actual off-station load D_m-1Conventional number of visitors X who do not go to large-scale activity place at downstream station in current direction_Dm-1:m-1+p，p>＝1；

b) Estimation of temporary passenger train C_mThe conventional number of persons who do not go to the large-scale event place in the load of the upstream station leaving the station and the number of persons who get off the large-scale event place in the part of conventional passenger flow at the respective downstream stations are as follows:

consider C_mThe number of regular persons who do not go to the large-scale activity place in the load of the upstream station leaving the station and the number of persons who get off the regular passenger flow at the respective downstream stations are both 0;

c) estimating the number of regular passengers who do not go to a large-scale activity place in the load of each station leaving after the operation of the train A and the upstream late train is recovered and the passenger train is inserted into the station and operated and the number of passengers who leave the station in the part of regular passenger flow at each downstream station:

according to the nearest punctual train at C_m-1+pActual number of passengers at station, C_m-1+pNumber of passengers on model station, current train in C_m-1+pEstimating current train at C by actual number of guests at station_m-1+pConventional number of guests S who do not go to large event place among the number of guests actually standing_{m-1+ p: general of}，p>＝1；

According to the conventional number of people who do not go to large-scale activity places in the load of leaving the station at the upstream station, C_m-1+pNumber of passengers getting off at station_m-1+p、C_m-1+pNumber of regular guests S estimated from number of actual guests in station_{m-1+ p: general of}Calculating C_m-1+pThe number of the conventional people in the large-scale activity place is not increased in the load of the station leaving the station;

according to S_{m-1+ p: general of}And the OD model calculates the train is in C_m-1+pNumber of regular guests S of station_{m-1+ p: general of}The number of passengers getting off at the downstream station X_{m-1+p:m-1+p+q}，p>＝1，q>＝1；

According to the number of passengers getting off the downstream station and X of the conventional number of people who do not go to the large-scale activity place in the load of the upstream station leaving the station_{m-1+p:m-1+p+q}Calculating train C_m-1+pLoad D from station to station_m-1+pRegular people who do not go to large-scale activity places can get off at downstream station by X number of passengers_{Dm-1+p:m-1+p+q}，p>＝1，q>＝1；

Train at C_m-1+pActual number of visitors, p>The train passenger flow detection system can detect passenger flow in a multivariate passenger flow detection mode such as train weighing and the like as 1.

Preferably, the step 1.2.4.2) is specifically:

a)R_j,j+1for train A at station C_jTo C_j+1Load of interval, j belongs to [ i-n, i-1 ]]The passenger flow is obtained by a multi-element passenger flow detection mode such as train weighing;

b) predicting train A at station C_kThe number of passengers leaving, the number of passengers getting on, the number of regular passengers not leaving the large activity place in the off-station load and the number of passengers leaving the regular passenger flow at the downstream station, k is [ j +1, i-1 ]]：

Train A at station C_kPredicting the number of passengers getting off X_kIs X_Dk-1:k；

Train A at station C_kPredicted number of guests S_kComprises the following steps:

b1) if the train A and the front train are at the station C_kDoes not meet the planned interval, then:

calculating and predicting the number of passengers getting on the train according to the ratio of the actual interval to the planned interval and the number of the passengers getting on the train at the punctual point where the train stops at the station;

train A is at C_kStation passenger number prediction getting-on vehicleThe smaller of the number of people I and the number of available guests of the train A;

b2) if the train A and the front train are at the station C_kThe arrival intervals of (a) meet the planned intervals, then:

can refer to the front vehicle at C_kThe number of passengers on the station is calculated to predict the number of passengers on the station; train A is at C_kThe number of the passengers on the station is the smaller value of the predicted number of the passengers on the train A and the number of the passengers on the train A;

according to the nearest punctual train at C_kActual number of passengers at station, C_kNumber of passengers on model station, current train in C_kStation prediction guest number estimation current train in C_kConventional number of guests S who do not go to large-scale event place among predicted number of guests at station S_{k: general of}；

According to the conventional number of people who do not go to large-scale activity places in the load of leaving the station at the upstream station, C_kStation forecast number of passengers X_k、C_kConventional number of guests S estimated from predicted number of guests at station_{k: general of}Calculating C_kThe number of the conventional people in the large-scale activity place is not increased in the load of the station leaving the station;

according to S_{k: general of}And the OD model calculates the train is in C_kNumber of regular guests S of station_{k: general of}The number of passengers getting off at the downstream station X_k:k+q，q>＝1；

According to the number of passengers getting off the downstream station and X of the conventional number of people who do not go to the large-scale activity place in the load of the upstream station leaving the station_k:k+qCalculating train C_kLoad D from station to station_kRegular people who do not go to large-scale activity places can get off at downstream station by X number of passengers_Dk:k+q，q>＝1；

c) Predicting train A at station C_iThe number of underground cars:

if train A participates in the calculation of step 1.2.4.1), train A can be referred to as C_i-1Off-station loading of a station D_i-1Estimated D_i-1The number of persons who can not go to large-scale activity places and the part of conventional passenger flow at station C_iThe number of passengers getting off the train A at the station C_iEstimating the number of getting-off people;

if train A is not involved in the calculation in step 1.2.4.1), then train A at C can be considered as a function of train A_iPredicted arrival load, model arrival load, and model number of passengers in train A at C_iThe number of passengers getting off.

d) Station C_i-nTo C_i-1Each train leaving the station of updating C_iThe number of passengers is predicted.

Preferably, the station passenger outbound capacity acquisition specifically includes:

2.1) detecting passenger flows at a platform, a key channel and an exit by adopting a multivariate passenger flow detection means at a station;

2.2) the station reports the outbound rate of the current bottleneck area periodically, and the buffer area at the upstream of the bottleneck area can also accommodate a plurality of people.

Preferably, the train operation capacity adjustment is specifically as follows:

3.1) to reduce the passenger quantity under the train at the station Ci, an upstream station can be selected for passenger clearing:

3.1.1) at C_i-nTo C_iSelecting 1 stations C with small passenger quantity from the stations in between_rClearing the visitor;

3.1.2) if station C_iIs a terminal station, if station C_rConnecting the train section or the parking lot, the train can go back to the train section or the parking lot for waiting after clearing passengers, and then goes out of the train section or the parking lot from the station C_rCurrent service number slave station C for train starting to run_iThe next task after the turn-back; if the station C_rNot connecting the train section or the parking lot, the train keeps going forward after clearing the passenger, and the train is at the station C_iContinuing to carry passengers for operation after turning back;

3.1.3) if station C_iIf the train is not a terminal station, the train keeps going forward after clearing passengers and is at a station C_iThe next station continues to carry passengers;

3.1.4) discontinuously clearing the passengers, and adopting the decision of setting the passenger clearing time;

3.2) to reduce station C_iIf the train leaves the passenger capacity of station C_iInstead of terminal stations, it is also possible to select a part of train stops C_i：

Station C_iReporting: station C_iHas a passenger outbound rate V in a bottleneck region_o-iThe upstream buffer of the bottleneck region can accommodate R at most_i:maxPeople, at present, can also accommodate passengers R_iA human;

station C_rPlatform to station C_iThe platform has k vehicles including ascending and descending, the last vehicle is train A and station C_rThe first train B in the upstream section still requires time T1 to arrive at the station C_iAfter time period T1, station C_iWait for departure people increment △ O_iIs (k trains are at C)_iStation forecast leaving passenger total-V_o-i*T1)，R_iTaking (R)_i-△O_i) And R_i:maxOf where V is_o-iIs station C_iThe passenger outbound rate in the bottleneck region;

if (R)_i+ train B at C_iPredicted number of passengers getting off at station) threshold F, train B needs to be at station C_iJumping and stopping;

if the second train upstream of train B is also at C_iWhen the station stops, if the train and the train B are trains in the same direction, the first train on the upstream of the train B is considered to stop so as to avoid continuous stop in the same direction;

whenever station C is updated_iWhen the number of passengers is predicted, jump stop decision is made.

Preferably, the 3.1.4) decision of the passenger clearing opportunity is specifically:

station C_r-1Platform to station C_iThe platform has k vehicles including up and down, the last vehicle is train A and station C_r-1The first train B in the upstream section still requires time T1 to arrive at the station C_iAfter time period T1, station C_iWait for departure people increment △ O_iIs (k trains are at C)_iStation forecast leaving passenger total-V_o-i*T1)，R_iTaking (R)_i-△O_i) And R_i:maxThe smaller of these;

if (R)_i+ train B at C_iPredicted number of passengers getting off at station) threshold F, train B needs to be at station C_rClearing away guests;

the number of the full passengers in the train is T_maxIf the train A is full of passengers and arrives at the station C_iAfter, R_iApproaching the threshold value F if the station C is_iCannot complete T within 2 train intervals_maxThe number of people going out, at which time if the train on the upstream of the train B arrives full of people, the train may also be cleared of passengers, if the station C_iIf the terminal station is used, the situation that two trains of passengers are cleared continuously can occur;

if the station C_iNot at the terminal station, but also at C if the second train upstream of train B_iWhen the train and the train B are trains in the same direction, the first train on the upstream of the train B is considered to avoid continuous passenger clearing in the same direction;

whenever station C is updated_iWhen the number of passengers is predicted, passenger clearing decisions are made.

Compared with the prior art, the invention has the following advantages:

1. according to the invention, big data analysis is carried out based on data such as ticket service clearing data and train weighing, and a train departure load, passenger number of getting-on and getting-off and OD passenger flow model of each station is established aiming at the current schedule, so that model support is provided for passenger flow prediction.

2. The invention can carry out real-time train leaving load detection and train getting-on and getting-off number detection at each station based on the multi-passenger flow detection modes such as train weighing and the like, and can carry out train getting-off amount prediction at a specified station in time.

3. The invention detects the passenger flow at the station platform, the key channel, the exit and the like based on a multivariate passenger flow detection means, and can know in real time how many people can be accommodated in the upstream buffer zone of the outbound bottleneck area.

4. The invention can scientifically adjust the transportation capacity according to the station passenger outbound capacity based on the information.

Drawings

FIG. 1 is a data relationship diagram of the present invention;

FIG. 2 is a circuit diagram of an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, shall fall within the scope of protection of the present invention.

A station route diagram of an embodiment of the invention is shown in fig. 2. An intelligent train dispatching method for urban rail transit passenger flow loop self-adaption comprises the steps of predicting the passenger quantity under a train, obtaining the station passenger outbound capacity and adjusting the train transport capacity.

The method for predicting the passenger capacity under the train comprises the following steps:

1) big data analysis is carried out based on data such as ticket service clearing data and train weighing, and a train leaving load, passenger number of getting-on and getting-off and OD passenger flow model of each station are established according to the current schedule.

2) If a certain station C_iExit linkage large-scale activity place is holding large-scale exhibition, motion meeting etc. in a few consecutive days, need carry out safety inspection to the personnel of the place of advancing of leaving a station to partial export has been closed according to the place receptivity, for reaching speed control with the passenger this moment in station passenger ability within range of leaving a station, need foresee the volume of getting off a train:

A. passenger arrival rate prediction preconditions: station C in the day_iExcept that there are no similar large activities near the remaining stations.

B. Under normal conditions, at C_iWithin the range of n upstream stations, if the arrival intervals of the trains and the front vehicle on the same platform accord with the planned interval and the arrival intervals of the trains and the front vehicle on the same platform on the upstream of the current direction also accord with the planned interval, the departure load of the trains and the arrival intervals of the trains on the C station can be determined according to the departure load of the trains_i-n+1To C_i-1On each station between stationsNumber of passengers getting off and passenger flow model for train C_iIf the number of passengers is predicted, if the station C_iNot a terminal station, the up and down trains need statistics:

a)R_j,j+1for train A at station C_jTo C_j+1Load of interval, j belongs to [ i-n, i-1 ]]And the passenger flow is obtained by a multi-element passenger flow detection mode such as train weighing.

b)X_kFor train A at station C_kThe number of passengers, k, belongs to [ j +1, i-1 ]]The model value or the number of passengers getting off the train at the station can be referred to.

c)S_kFor train A at station C_kThe number of passengers getting on, k belongs to [ j +1, i-1 ]]The predicted number of passengers can be calculated by referring to the model value or the number of the most recent passengers on the train which stops at the station. Train A is at C_kThe number of passengers on the station is the smaller value of the predicted number of passengers on the train A and the number of passengers on the train A.

d) According to train at C_iPredicted arrival load, model arrival load, and model number of passengers in train_iThe number of passengers getting off.

e) Train at C_kK is [ i-n +1, i ]]The system can detect passenger flow in multiple passenger flow detection modes such as train weighing.

f) Station C_i-nTo C_i-1Each train leaving the station of updating C_iThe number of passengers is predicted.

C. If C is present_jThe station has shielding door fault, j belongs to [ i-n +1, i-1]Train at C_jThe station adopts the over-station adjustment strategy, then C_jThe number of passengers getting on and off the station is 0, and the passengers on the vehicle are generally selected to be in the position C_j-1Station and C_j+1Getting off the bus, if j is i-1, the bus needs to be checked at the bus station C_iMake an announcement that the number of newly added visitors is likely to be in C_iWhen the train leaves the station, the train can be transferred to the reverse train to return to the station C_i-1And (4) a station.

D. If the failure treatment results in the threshold value F that the train A and the previous train are on the same platform (actual departure interval-planned departure interval), the upstream part of the train A is buckled and stopped at a late point, and the train A and the previous train are buckledThe part of the train waiting for the train A and the train waiting for the train waiting_iThe number of passengers getting off the elevator is specifically predicted. Let C be_mFor the station when the train returns to operation, the next train is the starting station for inserting operation, and m belongs to [ i-1, 1]：

a) Estimating the number of regular passengers who do not go to a large-scale activity place in the departure load when the train leaves the station and the train A and the train at the upstream late point and the number of passengers who get off the station in the regular passenger flow at the respective downstream station:

estimate train A and upstream late train C_mThe conventional number of persons who do not go to the large-scale event place in the load of the upstream station leaving the station and the number of persons who get off the large-scale event place in the part of conventional passenger flow at the respective downstream stations are as follows:

conventional passenger flow OD model of current time period and current direction without going to large-scale activity place: at station C_xAmong passengers getting on will have B_x:x+y% of passengers in C_x+yStanding passenger, y>＝1。

According to the OD model, the train is at the station C_mModel number of guests on each station at upstream calculation station C_m-1Model off-station load D_m-1Number of visitors under model of downstream station in current direction X_Dm-1:m-1+p，p>＝1。

Suppose that the current train is at station C_mThe number of guests at each upstream station includes the number of specific guests going to the large event and the number of model guests not going to the large event, and C_m-1Model off-station load D_m-1Number of visitors under model of downstream station in current direction X_Dm-1:m-1+pAs the current train C_m-1Actual off-station load D_m-1Conventional number of visitors X who do not go to large-scale activity place at downstream station in current direction_Dm-1:m-1+p，p>＝1。

Estimate oncoming traffic C_mThe conventional number of persons who do not go to the large-scale event place in the load of the upstream station leaving the station and the number of persons who get off the large-scale event place in the part of conventional passenger flow at the respective downstream stations are as follows: the passenger car is generally empty, so C can be considered as_mRegular people who do not go to large-scale activity places in load of leaving station at upstream stationThe number of passengers in the conventional passenger flow at the downstream stations is 0.

Estimating the number of regular passengers who do not go to a large-scale activity place in the load of each station leaving after the train A and the train at the upstream late point are recovered to operate and the passenger flow of the regular passengers at each downstream station after the adjacent passenger train is inserted into the station to operate:

according to the nearest punctual train at C_m-1+pActual number of passengers at station, C_m-1+pNumber of passengers on model station, current train in C_m-1+pEstimating current train at C by actual number of guests at station_m-1+pConventional number of guests S who do not go to large event place among the number of guests actually standing_{m-1+ p: general of}，p>＝1。

According to the conventional number of people who do not go to large-scale activity places in the load of leaving the station at the upstream station, C_m-1+pNumber of passengers getting off at station_m-1+p、C_m-1+pNumber of regular guests S estimated from number of actual guests in station_{m-1+ p: general of}Calculating C_m-1+pThe number of persons who normally go to large activity places in the load of leaving the station is not limited.

According to S_{m-1+ p: general of}And the OD model calculates the train is in C_m-1+pNumber of regular guests S of station_{m-1+ p: general of}The number of passengers getting off at the downstream station X_{m-1+p:m-1+p+q}，p>＝1，q>＝1。

According to the number of passengers getting off the downstream station and X of the conventional number of people who do not go to the large-scale activity place in the load of the upstream station leaving the station_{m-1+p:m-1+p+q}Calculating train C_m-1+pLoad D from station to station_m-1+pRegular people who do not go to large-scale activity places can get off at downstream station by X number of passengers_{Dm-1+p:m-1+p+q}，p>＝1，q>＝1。

b) Prediction C_iTrain in the range of upstream n stations is at C_iThe number of passengers getting off, if the station C_iNot a terminal station, the up and down trains need statistics:

·R_j,j+1for train A at station C_jTo C_j+1Load of interval, j belongs to [ i-n, i-1 ]]And the passenger flow is obtained by a multi-element passenger flow detection mode such as train weighing.

Prediction of train A at station C_kThe number of passengers leaving, the number of passengers getting on, the number of regular passengers not leaving the large activity place in the off-station load and the number of passengers leaving the regular passenger flow at the downstream station, k is [ j +1, i-1 ]]：

Train A at station C_kPredicting the number of passengers getting off X_kIs X_Dk-1:k。

if the train A and the front train are at the station C_kDoes not meet the planned interval, then:

and calculating the predicted getting-on number I according to the ratio of the actual interval to the planned interval and the number of the last guests of the quasi-point train at the stop.

Train A is at C_kThe number of passengers on the station is the smaller value of the predicted number of passengers on the train A and the number of passengers on the train A.

If the train A and the front train are at the station C_kThe arrival intervals of (a) meet the planned intervals, then:

can refer to the front vehicle at C_kThe number of passengers on the station is calculated to predict the number of passengers on the station, namely the number of passengers on the station is one, and the number of front vehicles is C_kThe number of guests at a station may be an actual value or a predicted value.

According to the nearest punctual train at C_kActual number of passengers at station, C_kNumber of passengers on model station, current train in C_kStation prediction guest number estimation current train in C_kConventional number of guests S who do not go to large-scale event place among predicted number of guests at station S_{k: general of}。

According to the conventional number of people who do not go to large-scale activity places in the load of leaving the station at the upstream station, C_kStation forecast number of passengers X_k、C_kRoutine number of guests estimated from predicted number of guestsS_{k: general of}Calculating C_kThe number of persons who normally go to large activity places in the load of leaving the station is not limited.

According to S_{k: general of}And the OD model calculates the train is in C_kNumber of regular guests S of station_k：_{General of}The number of passengers getting off at the downstream station X_k:k+q，q>＝1。

According to the number of passengers getting off the downstream station and X of the conventional number of people who do not go to the large-scale activity place in the load of the upstream station leaving the station_k:k+qCalculating train C_kLoad D from station to station_kRegular people who do not go to large-scale activity places can get off at downstream station by X number of passengers_Dk:k+q，q>＝1。

Prediction of train A at station C_iThe number of underground cars:

if train A participates in the calculation in a), reference can be made to train A at C_i-1Off-station loading of a station D_i-1Estimated D_i-1The number of persons who can not go to large-scale activity places and the part of conventional passenger flow at station C_iThe number of passengers getting off the train A at the station C_iEstimating the number of alighting persons.

If train A does not participate in the calculation in a), it can be determined from train A at C_iPredicted arrival load, model arrival load, and model number of passengers in train A at C_iThe number of passengers getting off.

Station C_i-nTo C_i-1Each train leaving the station of updating C_iThe number of passengers is predicted. The method for acquiring the outbound capacity of the passenger in the station comprises the following steps:

1) the station adopts a multi-element passenger flow detection means to detect passenger flows at platforms, key channels, exits and the like.

2) The station periodically reports the outbound rate of the current bottleneck area, and the buffer area at the upstream of the bottleneck area can also accommodate a plurality of people.

The train capacity adjusting method comprises the following steps:

1) for reducing station C_iThe passenger capacity under the train can be selected to clear passengers at an upstream station:

A. at C_i-nTo C_iSelect 1 in station betweenStation C with small passenger capacity_rAnd clearing the passengers to avoid the conflict of passenger flow caused by clearing the passengers.

B. If the station C_iIs a terminal station, if station C_rConnecting the train section/parking lot, the train can go back to the train section/parking lot for waiting after clearing passengers, and then goes out of the train section/parking lot from the station C_rCurrent service number slave station C for train starting to run_iThe next task after the turn-back; if the station C_rNot connecting the vehicle section/parking lot, the train keeps going forward after clearing the passenger, and at the station C_iAnd continuing to carry passengers for operation after turning back.

C. If the station C_iIf the train is not a terminal station, the train keeps going forward after clearing passengers and is at a station C_iThe next station continues to carry passengers.

D. The customer is generally not cleared continuously.

E. And (3) deciding the opportunity of clearing the visitor:

a) station C_iReporting: station C_iHas a passenger outbound rate V in a bottleneck region_o-iThe upstream buffer of the bottleneck region can accommodate R at most_i:maxPeople, at present, can also accommodate passengers R_iA human.

b) Station C_r-1Platform to station C_iThe platform has k vehicles including up and down, the last vehicle is train A and station C_r-1The first train B in the upstream section still requires time T1 to arrive at the station C_iAfter time period T1, station C_iWait for departure people increment △ O_iIs (k trains are at C)_iStation forecast leaving passenger total-V_o-i*T1)，R_iTaking (R)_i-△O_i) And R_i:maxThe smaller of these.

c) If (R)_i+ train B at C_iPredicted number of passengers getting off at station) threshold F, train B needs to be at station C_rClearing away the visitor.

d) The number of the full passengers in the train is T_maxIf the train A is full of passengers and arrives at the station C_iAfter, R_iApproaching the threshold value F if the station C is_iCannot complete T within 2 train intervals_maxThe number of people is out of the station, and if the upstream train of the train B is full of people to arrive,the train may also need to be cleared if station C_iThe terminal station can be used for continuously clearing passengers for two trains.

e) If the station C_iNot at the terminal station, but also at C if the second train upstream of train B_iWhen the train is clear and the train B are trains in the same direction, the first train on the upstream of the train B is considered to avoid the continuous clearing of passengers in the same direction.

f) Whenever station C is updated_iWhen the number of passengers is predicted, passenger clearing decisions are made.

2) For reducing station C_iIf the train leaves the passenger capacity of station C_iInstead of terminal stations, it is also possible to select a part of train stops C_i：

A. Station C_iReporting: station C_iHas a passenger outbound rate V in a bottleneck region_o-iThe upstream buffer of the bottleneck region can accommodate R at most_i:maxPeople, at present, can also accommodate passengers R_iA human.

B. Station C_rPlatform to station C_iThe platform has k vehicles including up and down, the last vehicle is train A and station C_rThe first train B in the upstream section still requires time T1 to arrive at the station C_iAfter time period T1, station C_iWait for departure people increment △ O_iIs (k trains are at C)_iStation forecast leaving passenger total-V_o-i*T1)，R_iTaking (R)_i-△O_i) And R_i:maxThe smaller of these.

C. If (R)_i+ train B at C_iPredicted number of passengers getting off at station) threshold F, train B needs to be at station C_iAnd (6) stopping jumping.

D. If the second train upstream of train B is also at C_iAnd when the station stops, and the train B are trains in the same direction, considering the first train on the upstream of the train B to stop so as to avoid continuous stop in the same direction.

E. Whenever station C is updated_iWhen the number of passengers is predicted, jump stop decision is made.

While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and various equivalent modifications and substitutions can be easily made by those skilled in the art within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. An intelligent train dispatching method for urban rail transit passenger flow loop self-adaption is characterized by comprising the following steps:

step 1), predicting the off-train passenger capacity;

step 2), acquiring the outbound capacity of passengers at a station;

2. The method for intelligent adaptive train dispatching in urban rail transit passenger flow loop according to claim 1, wherein the under-train passenger volume prediction specifically comprises:

step 1.1), carrying out big data analysis based on ticket service clearing data and train weighing data, and establishing an OD model of train departure load, passenger number of getting-on and getting-off and passenger flow of each station aiming at a current schedule;

step 1.2), if a certain station C_iThe exit is connected with a large-scale event place, large-scale events are held within a plurality of days continuously, safety check needs to be carried out on personnel going out of the station and entering the place, partial exits are closed according to the place reception capacity, and at the moment, the passenger getting-off amount needs to be predicted in order to control the passenger arrival rate within the passenger going-out capacity range of the station.

3. The urban rail transit passenger flow loop self-adaptive intelligent train dispatching method according to claim 2, wherein the predicting of the passenger volume under the train specifically comprises:

step 1.2.4) if the actual departure interval-planned departure interval of the train A and the previous train on the same platform is larger than the threshold value F due to fault treatment, the upstream part of the train A is buckled and stopped at a late point, a part of temporary trains are inserted between the train A and the previous train for reducing the large interval, the passenger carrying conditions of the temporary trains, the train A and the upstream late point train are different from those of the normal alignment point train, and the part of the train is required to be positioned at a station C_iThe number of passengers getting off the elevator is specifically predicted.

4. The urban rail transit passenger flow loop self-adaptive intelligent train dispatching method according to claim 3, wherein the people counting in step 1.2.2) is specifically as follows:

step 1.2.2.1) R_j,j+1For train A at station C_jTo C_j+1Load of interval, j belongs to [ i-n, i-1 ]]Obtaining the passenger flow through a multi-element passenger flow detection mode;

step 1.2.2.4) according to train on C_iPredicted arrival load, model arrival load, and model number of passengers in train_iThe number of passengers getting off, the train is at C_kK is [ i-n +1, i ]]The detection can be carried out by a multivariate passenger flow detection mode;

5. The method for intelligent city rail transit passenger flow loop adaptive train dispatching according to claim 3, wherein the specific prediction in the step 1.2.4) is as follows:

6. The urban rail transit passenger flow loop adaptive intelligent train dispatching method according to claim 5, wherein the step 1.2.4.1) specifically comprises:

a) estimating train A and upstream late train C_mThe conventional number of people who do not go to large-scale activity places in the load of the upstream station leaving the station and the conventional passenger flow of the part are respectively at the downstream stationsThe number of passengers:

conventional passenger flow OD model of current time period and current direction without going to large-scale activity place: at station C_xAmong passengers getting on will have B_x:x+y% of passengers in C_x+yStanding passenger, y>＝1；

According to the condition that large-scale activities are not removed in the load of leaving the station at the upstream stationGeneral number of people in moving place, C_m-1+pNumber of passengers getting off at station_m-1+p、C_m-1+pNumber of regular guests S estimated from number of actual guests in station_{m-1+ p: general of}Calculating C_m-1+pThe number of the conventional people in the large-scale activity place is not increased in the load of the station leaving the station;

Train at C_m-1+pActual number of visitors, p>The detection can be carried out by a multivariate passenger flow detection mode as 1.

7. The urban rail transit passenger flow loop adaptive intelligent train dispatching method according to claim 5, wherein the step 1.2.4.2) specifically comprises:

a)R_j,j+1for train A at station C_jTo C_j+1Load of interval, j belongs to [ i-n, i-1 ]]Obtaining the passenger flow through a multi-element passenger flow detection mode;

train A is at C_kThe number of the passengers on the station is the smaller value of the predicted number of the passengers on the train A and the number of the passengers on the train A;

c) Predicting train A at station C_iThe number of underground cars:

8. The urban rail transit passenger flow loop self-adaptive intelligent train dispatching method according to claim 1, wherein the station passenger outbound capacity acquisition specifically comprises:

9. The urban rail transit passenger flow loop adaptive intelligent train dispatching method according to claim 1, wherein the train capacity adjustment specifically comprises:

3.1.2) if station C_iIs a terminal station, if station C_rConnecting the train section or the parking lot, the train can go back to the train section or the parking lot for waiting after clearing passengers, and then goes out of the train section or the parking lot from the station C_rCurrent service number slave station C for train starting to run_iThe next task after the turn-back; if the station C_rNot connecting the train section or the parking lot, the train keeps going forward after clearing the passenger, and the train is at the station C_iContinue to carry after turning backThe guest runs;

10. The urban rail transit passenger flow loop adaptive intelligent train scheduling method according to claim 9, wherein the 3.1.4) passenger clearing opportunity decision is specifically: