TWI658414B - Vehicles dispatch system and method - Google Patents

Abstract

The invention discloses a vehicle dispatching system and a method thereof. This system and method includes establishing a database of boarding probability statistics based on the relationship between weather factors and specific places, a computer system for passenger demand, and a passenger demand correction mechanism. Through the present invention, the passenger's forecasted demand can be calculated based on the statistical data base of the boarding probability, combined with the location and weather parameters, and then the passenger's demand can be corrected in real time based on the actual passenger's demand reported by the driver. By continuously accumulating the passenger probability statistics database of the relationship between passenger demand and weather, and using the computerized system of passenger demand forecast and the real-time correction mechanism of passenger demand, a more efficient optimization method for estimating passenger demand is realized. .

Description

Vehicle dispatch system and method

The present invention relates to a vehicle dispatching technology, and more particularly, to a vehicle dispatching system and method for predicting and real-time correcting passenger demand in specific places based on weather factors.

Taxi is a land-based public transport service based on instant rental. Generally speaking, taxis usually charge according to the odometer. In addition to calling at taxis, passengers can also make appointments by phone or online.

That is to say, in addition to taxis going to a fixed number of people to stop passengers on the roadside, another taxi calling service usually requires the user to call the taxi's management center, which then chooses through different mechanisms. And dispatch a specific taxi to the place designated by the user.

However, although the current taxi operation dispatch management system has a mechanism for matching taxis with dispatch media, it can only passively wait for the demand of passengers to call for a taxi, and then provide a matchmaking service for drivers and passengers.

If there are too many taxis in a place with a large number of residents in a single area (such as a station, hospital, or department store), it often happens that the schedule is too long, and the number of empty taxis is too high, the load factor is low, or the number of taxis is high. Customer situation As a result, the operating costs of taxis have increased and revenue has been reduced, which is quite inconsistent with economic benefits. Or, there are other factors such as the weather, which also depends on the willingness of the passengers to carry them, and the management center of the derived taxis has uncertainties in the number of dispatched taxis.

It can be seen that there are still many shortcomings in the above-mentioned customary methods. It is not a good design, and it needs to be improved. In view of various shortcomings derived from the above-mentioned conventional methods, the inventor of the present invention has been eager to improve and innovate. After painstaking research, he finally successfully developed the vehicle dispatching system and method of the present invention.

The invention provides a vehicle dispatching system and a method thereof, which can perform passenger demand prediction and real-time correction of specific places according to weather factors. First, the passenger demand system obtains the real-time resident number and the current weather factor (such as the probability of rainfall, etc.) of a specific place for a specific place, and uses this two factors (location and weather) to input the boarding probability statistics database to obtain this. Under the conditions of the boarding probability K factor, the passenger demand is calculated by multiplying the number of real-time residents by the boarding probability K factor. Secondly, a considerable proportion of vehicles to the predicted demand of passengers are dispatched to the site. Then, based on the actual passenger demand value reported by the driver arriving at the venue to the passenger demand system, based on the relationship between the actual passenger demand and the passenger's forecasted demand, decide whether to dispatch a vehicle to the venue again, and recalculate the new boarding probability K Coefficient, and then based on the parameters of the two factors (location and weather) as conditions, the value of the corresponding K coefficient of the riding probability in the statistical database of riding probability is corrected in real time. Through this system and method, continuously calculate passenger demand for each specific place, and continuously accumulate passenger demand and days A statistical database of passenger probability based on air parameter relationships, combined with the computerized system of passenger demand forecasting and the real-time correction mechanism of passenger demand, in order to achieve a more efficient estimation of passenger demand optimization.

According to the above object, the present invention provides a vehicle dispatching system, including: a number of people demand module, counting the number of people staying at a place within a predetermined time; and a module of capturing the number of people staying within the predetermined time. The current latest resident information S _{A of the place} ; a weather data module to obtain the current one or more weather parameters R of the place; a boarding probability calculation module based on the weather parameters R and the place parameters P of the place calculate the probability K _a ride aboard the chances of statistics from the database; a number of demand calculation module to calculate a forecast passenger demand N _a number of people according to the latest information resides K _a S _a ride with the probability; and a dispatch mode group, the passenger demand prediction based on N _A to the property sent to a corresponding number of vehicles. According to the above object, the present invention further provides a vehicle dispatching method, which includes the following steps: acquiring a latest resident number information S _{A of} a current place within a predetermined time through an acquisition resident number module, and calculating by a boarding probability The module calculates a boarding probability K _A from the boarding probability statistics database according to one or more weather parameters R and the place parameters P of the place; through a person demand calculation module, the module is based on the latest resident information S _A and the boarding The probability K _A calculates a predicted passenger demand N _A ; and dispatches a corresponding number of vehicles to the site according to the predicted demand N _{A of} the passenger through a dispatch module.

Therefore, the technical advantages of the present invention are as follows.

In the system proposed by the present invention, the passenger demand is estimated by combining the number of people in a specific place and weather factors, and the boarding probability is corrected in real time based on the actual passenger load.

The present invention focuses on pre-calculating passenger demand information for the fleet to carry out passenger carrying planning in advance, and to estimate potential passenger carrying business opportunities.

The invention only needs the number of people in a specific place and the weather to perform calculations. In addition to reducing the amount of data and calculations, the number of passengers required can be more accurately estimated.

The invention also proposes a real-time correction mechanism for the actual passenger load, which adjusts the passenger load demand in real time, so that fleet management can more accurately control the number of passengers.

10‧‧‧ Vehicle Dispatching System

100‧‧‧ Passenger Demand System

110‧‧‧Timer Demand Module

120‧‧‧Ride probability calculation module

130‧‧‧Weather Information Module

140‧‧‧Retrieve resident module

150‧‧‧person demand calculation module

160‧‧‧Dispatching module

170‧‧‧Riding probability K coefficient correction module

180‧‧‧Ride Probability Statistics Database

200‧‧‧ Place

210‧‧‧Real-time resident module

300‧‧‧ Vehicle

310‧‧‧ Passenger Actual Demand Reporting Module

C _A ‧‧‧ actual passenger demand

K‧‧‧ Probability

K _A ‧‧‧ Probability

K'‧‧‧ modified boarding probability

N _A ‧‧‧ Passenger Forecast Demand

P (P _A to P _G ) ‧‧‧Place parameters

R‧‧‧ Weather Parameters

S _A ‧‧‧ latest resident information

S201 ~ 206, S401 ~ 409‧‧‧ steps

T _A ‧‧‧ actual passenger boarding probability

The specific embodiments disclosed in the present invention will be described in detail with the following drawings, and these descriptions are shown in the following drawings: FIG. 1 is a schematic architecture diagram of a vehicle dispatch system of the present invention; FIG. 2 is one of the vehicle dispatch methods of the present invention Schematic flowchart; Figure 3 is a schematic diagram of the historical data table of the boarding probability K coefficient of the present invention; and Figure 4 is another schematic flowchart of the vehicle dispatch method of the present invention.

The preferred embodiment of the present invention is a passenger demand system applied to a taxi The purpose of advance vehicle dispatch by fleet operators is explained. In other words, the fleet operator can use the passenger demand system to target a specific place, in accordance with weather parameters, to estimate the passenger demand in advance and dispatch a considerable proportion of the estimated amount of vehicles to the specific place, in order to obtain subsequent passenger business opportunities. In addition, the driver can report the actual passenger demand of the venue to the passenger demand system, and then perform subsequent dispatch and correction of immediate demand.

Figure 1 is a schematic architecture diagram of the vehicle dispatch system 10 of the present invention. The entire system is composed of a passenger demand system 100 and a place 200 (such as a station, department store, movie theater, stadium, concert hall, theater, exhibition hall, etc.) With multiple vehicles 300. This embodiment uses a station as an example. The vehicle dispatching system 10 includes a time counting demand module 110, a boarding probability calculation module 120, a weather data module 130, a capture resident module 140, a crowd demand calculation module 150, a dispatch module 160, and a boarding probability K Coefficient correction module 170 and boarding probability statistics database 180. The site 200 includes a real-time resident number module 210, which provides the real-time resident number data to the passenger demand system 100 as a reference source of demand statistics. The vehicle 300 includes a passenger actual demand return module 310, which is returned to the passenger demand system 100 according to the actual number of passengers after the driver arrives at the location 200 for the purpose of dispatching and revising the K-factor of the boarding probability.

When referring to the schematic architecture diagram of the vehicle dispatch system 10 in FIG. 1, please also refer to one schematic flowchart of the vehicle dispatch method of the present invention shown in FIG. 2.

In the implementation step of calculating passenger demand, the passenger demand forecast calculation for each specific place 200 is started by the time-counting passenger demand module 110 move. The real-time resident module 210 in the place 200 (such as a station) records the latest resident number of the place and returns the time-counted person demand module 110.

In step S201, the number of people counting module 110 can be timed in a fixed or irregular time interval or in an event-driven manner within a predetermined time. Taking the passenger demand calculation for the site 200 as an example, the operation steps of the passenger demand computer system will be described.

In step S202, when the time-counting-demand module 110 has timed out, the current weather parameter R of the place 200 is obtained through the weather data module 130 immediately. The weather parameter R may be one or a combination of information on rainfall probability, temperature, humidity, and ultraviolet. However, the order of selecting weather parameters can be selected according to the order of rainfall probability, temperature, humidity, and ultraviolet, that is, the rainfall probability is preferentially selected when the chance of rainfall is high, and the selection is based on the factors of temperature, humidity, and ultraviolet.

In step S203, the boarding probability calculation module 120 calculates a boarding probability K _A from the historical boarding probability statistical database 180 under the conditions of the place parameter P and the weather parameter R, as shown in the following formula ( 1) shown.

K _A = K (P, R) Formula (1)

In step S204, the current resident number parameter S _{A of the} current place 200 is obtained from the instant resident number module 210 of the place 200 by retrieving the resident number module 140. Then, N _A passenger demand forecast according to the formula (2) of the passenger demand of calculation spaces 200.

N _A = S _A * K _A Formula (2)

In step S205, the taxi industry to follow the predicted demand accordingly passenger scheduling N _A considerable proportion of the estimated number of the vehicle to a place 200, the computer system described above is the sum of the prediction of passenger demand. The real-time correction mechanism for passenger demand begins to operate as the driver arrives at the venue 200. The driver can report to the passenger demand system 100 through the actual passenger demand report module 310 in the vehicle 300 according to the actual passenger demand C _A at the site 200, and the dispatch module 160 can predict the actual passenger demand C _A and the passenger's forecast. demand N relationship between _a and then decide whether to send more vehicles to place 200.

In step S206, the boarding probability K coefficient correction module 170 also performs correction calculation of the boarding probability K 'according to the following formula (3) and formula (4). First, formula (3) calculates the actual passenger boarding probability T _A. Secondly, the weighting calculation is performed according to the actual passenger riding probability T _A and the predicted riding probability K _A in formula (4) to calculate the modified riding probability K ′, where α is a coefficient between 0 and 1. Then, the revised boarding probability K ′ is updated to the boarding probability statistics database 180 according to the two factors of the location parameter P and the weather parameter R.

T _A = K _A * (C _A / N _A ) Formula (3)

K '= αT _A + (1-α) K _A Formula (4)

FIG. 3 is a schematic diagram of a historical data table of the boarding probability K coefficient of the present invention. This data table is stored in the boarding probability statistical database 180. The boarding probability K coefficient stored in the data table represents the boarding probability value under the conditions of the location 200 and the weather parameter R. The calculation method is as shown in the above formula (1). That is, the boarding probability statistics database 180 stores the corresponding boarding probability K _A according to the location 200 and the selected weather parameters R. In this table, the properties of parameter spaces 200 places parameters may be, for example, properties of the station P _A, a department store places the parameter P _B, movie theaters properties parameter P _C, the stadium property parameters P _D, P _E Concert Hall parameter , The venue parameter P _{F of the} theater and the venue parameter P _{G of the} exhibition hall, etc., but not limited to this.

This boarding probability K coefficient can be used in the passenger demand computer system to calculate the predicted passenger demand, as shown in the above formula (2). In addition, the passenger demand correction mechanism can also correct the boarding probability K factor for the error between the actual passenger demand and the passenger's predicted demand, as shown in the above formulas (3) and (4).

FIG. 4 is another schematic flowchart of the vehicle dispatching method of the present invention, which is a schematic flowchart of a system and method for forecasting and real-time correction of passenger demand at a specific location 200 based on weather factors (divided into the following two stages) Operation), and FIG. 1 shows a schematic architecture diagram of the vehicle dispatching system 10 of the present invention.

In step S401, within a predetermined period of time, the number-of-times demand module 110 may perform timing at a fixed or irregular time interval or an event-driven manner. The first stage is the operation of the passenger demand prediction mechanism. After the module 110 arrives, the demand calculation operation is started.

In step S402, the latest resident number information is obtained from the real-time resident number module 210 of the place 200 through the acquisition resident number module 140.

In step S403, the boarding probability calculation module 120 and the weather data module 130 bring the location parameter P and the weather parameter R, and use the above formula (1) in the boarding probability statistics database 180 to calculate the boarding probability K _A.

In step S404, the passenger demand calculation module 150 calculates the predicted passenger demand N _A of the place through the above formula (2), and then dispatches the number of vehicles corresponding to the predicted passenger demand to the place 200 through the dispatch module 160. This is the first stage of work.

In step S405, the second stage is the operation of the passenger demand correction mechanism. When the driver arrives at the venue 200, the actual passenger demand can be reported back to the passenger demand system 100 through the passenger actual demand return module 310 of the vehicle 300.

In step S406, the dispatch module 160 determines whether to dispatch more vehicles 300 to the site 200 based on the magnitude relationship between the actual passenger demand and the predicted passenger demand.

In step S407, if the actual passenger demand is greater than the predicted passenger demand, the dispatch module 160 dispatches more vehicles 300 to a specific place 200.

In step S408, if the actual passenger demand is less than or equal to the predicted passenger demand, the dispatch module 160 no longer dispatches more vehicles 300 to the specific place 200.

In step S409, after performing steps S407 and S408, the vehicle dispatching system 10 calculates the boarding probability K ′ after the boarding probability K coefficient correction module 170 is modified according to the above formula (3) and the above formula (4), Based on the two factors of the location parameter P and the weather parameter R, the modified boarding probability K 'is stored in the boarding probability statistical database 180 to facilitate the operation of the passenger demand prediction mechanism next time.

Then, according to the passenger demand forecasting mechanism and the passenger demand correction mechanism, this cycle of operations continues to be combined with the continuous accumulation and correction of the passenger probability statistics database to achieve a more efficient optimization of the estimated passenger demand. of.

The vehicle dispatching system and method of the present invention are described below by way of example. Please refer to Figs. 1 to 4 above.

First, calculate the predicted demand of passengers: P places the parameter (e.g., P _A) as a place (e.g., station) of, R is the weather or the weather parameter (hereinafter, the first consideration is the probability of rain factors), V ₈ is 0.8, S _A is the latest resident information (such as 200 people at the station). From the historical boarding probability statistics database 180, the boarding probability K _{A is} obtained according to the above formula (1), K _A = K (P _A , 0.8) = 0.7, and the predicted passenger demand N _A is calculated using the above formula (2), N _A = S _A * K _A = 200 * 0.7 = 140, so the predicted passenger demand N _A is 140 people.

2. Calculation of the number of dispatched vehicles: The vehicle dispatch system dispatches 50% of the number of vehicles to the place (such as a station). Assuming that a vehicle can carry 2 people, the vehicle is (140/2) * 50% = 35.

3. Calculation of the actual boarding probability: When the driver arrives at a place (such as a station), the actual passenger demand C _A is 160, and the actual passenger boarding probability T _{A is} calculated using the above formula (3), that is, T _A = K _A * ( C _A / N _A ) = 0.7 * (160/140) = 0.8.

Fourth, the real-time correction of passenger demand: According to the above formula (4), the actual passenger riding probability T _A and the predicted riding probability K _A are weighted to calculate the revised riding probability K ′, where α is 0.5, K '= αT _A + (1-α) K _A = 0.5 * 0.8 + 0.5 * 0.7 = 0.75, update the value of K' (0.75) to K (P _A , 0.8) in the boarding probability statistics database.

Compared with the current taxi operation and dispatch management system, although it has a mechanism for calling and dispatching media, it can only passively wait for the demand of passengers to call for a car, and then perform a matchmaking service for drivers and passengers. System and method for passenger demand forecasting and real-time correction in fixed places. Therefore, the present invention proposes a vehicle dispatching system and method for estimating and real-time passenger demand based on a statistical database of the probability of boarding, the passenger demand computer system, and the real-time passenger demand correction mechanism based on the relationship between weather factors and specific places. Amended.

The vehicle dispatching system and method of the present invention can provide an estimated passenger demand through a computer system that continuously accumulates the passenger demand and weather relationship and a real-time correction mechanism of the passenger demand. Optimization method of quantity. Therefore, the present invention has the following advantages: 1. Because accurate information on the number of resident in a specific place can be used as a reference source for passenger demand estimation, accurate reference of passenger source can be obtained; 2. As early as possible based on passenger demand forecast Configuring the vehicle in a specific place can save the waiting time of passengers, so it can shorten the waiting time of passengers; 3. It can increase the passenger demand business opportunities due to the weather factors, so it can increase the load factor; and 4. According to the actual passenger return by the driver Instant demand correction to improve passenger load efficiency, so demand can be corrected immediately.

The above-mentioned embodiment merely illustrates the principle, characteristics, and effects of this disclosure by way of example, and is not intended to limit the scope of implementation of this disclosure. Anyone who is familiar with this technique can do the above without departing from the spirit and scope of this disclosure. Modifications and changes to the implementation form. Any equivalent changes and modifications made by using the contents disclosed in this disclosure shall still be covered by the scope of patent application. Therefore, the scope of protection of the rights in this disclosure should be as listed in the scope of patent application.

Claims (13)

A vehicle dispatching system includes: a counting number and demand module, which counts the number of people staying in a place within a predetermined time; and an acquisition number module, which retrieves the latest information of the current number of people in the place within the predetermined time. S _A ; a weather data module that obtains the current one or more weather parameters R of the place; a boarding probability calculation module that calculates from the boarding probability statistics database based on the weather parameters R and the place parameters P of the place take out a probability K _a; a number of demand calculation module calculating a predicted passenger demand according to the latest number N _a K _a resident information S _a riding the probability, wherein the passenger demand forecast N _a is N _A = S _A * K _A ; and a dispatch module that dispatches a corresponding number of vehicles to the site based on the passenger's predicted demand N _A. The vehicle dispatching system according to item 1 of the scope of patent application, wherein the predetermined time is a fixed or non-fixed time interval. The vehicle dispatch system according to item 1 of the scope of the patent application, wherein the weather parameters R include one of a rainfall probability, a temperature, a humidity, and an ultraviolet ray or a combination thereof. The vehicle dispatching system described in item 1 of the scope of patent application, wherein the boarding probability K _A is K _A = K (P, R). The vehicle dispatch system described in item 1 of the scope of patent application, further includes a boarding probability K-factor correction module. When the dispatch module decides to continue to dispatch vehicles to the place, the boarding probability K-factor correction module is based on the boarding The probability K _A , an actual passenger demand C _A and the predicted demand N _A of the passenger are used to calculate an actual passenger boarding probability T _A based on the parameter α associated with the boarding probability K _A and the actual passenger boarding probability T _A A weight calculation is performed to calculate a revised boarding probability K ', and the boarding probability K stored in the boarding probability statistics database is updated according to the revised boarding probability K', where the α is between 0 Factor to 1. The vehicle dispatching system as described in item 5 of the scope of patent application, wherein the actual passenger boarding probability T _A is T _A = K _A * (C _A / N _A ), and the revised boarding probability K 'is K' = αT _A + (1-α) K _A. The vehicle dispatching system described in item 5 of the scope of the patent application, further includes a passenger actual demand return module for returning the actual number of passengers at the place to the boarding probability K-factor correction module as the revised Basis for boarding probability K '. The vehicle dispatching system described in item 1 of the scope of the patent application, wherein the boarding probability statistics database stores the corresponding boarding probability K _A according to the location and the screened weather parameters. The vehicle dispatching system described in item 1 of the scope of the patent application, further includes a real-time resident number module, which is used to record the latest resident number of the place and return it to the timed number-of-persons module. A vehicle dispatching method includes the following steps: Retrieving the latest resident number information S _{A of} a current place within a predetermined time through a retrieving resident number module, and using a boarding probability calculation module based on one or more weather parameters R and the place parameter P of the place calculate a boarding probability K _A from the boarding probability statistics database; a passenger demand calculation module calculates a passenger forecast based on the latest resident information S _A and the boarding probability K _A demand N _a, wherein, the passenger demand forecast is N _{a N a = S a * K a} ; and a sending module through the passenger demand prediction based on N _a corresponding to the number of vehicles dispatched to the location. The vehicle dispatch method as described in item 10 of the scope of patent application, wherein the boarding probability K _A is K _A = K (P, R). The vehicle dispatch method as described in item 10 of the scope of patent application, wherein when the dispatch module decides to continue dispatching vehicles to the place, a boarding probability K coefficient correction module is used according to the boarding probability K _A The amount C _A and the predicted demand N _A of the passenger are used to calculate an actual passenger boarding probability T _A , and a weight calculation is performed based on the parameter α associated with the boarding probability K _A and the actual passenger boarding probability T _A to calculate a correction. The subsequent boarding probability K ′ is further updated according to the modified boarding probability K ′, where the α is a coefficient between 0 and 1. The vehicle dispatch method according to item 12 of the scope of patent application, wherein the actual passenger boarding probability T _A is T _A = K _A * (C _A / N _A ), and the revised boarding probability K 'is K' = αT _A + (1-α) K _A.