KR101401787B1 - Method and apparatus for a tour recommendation service for electric vehicles based on a hybrid orienteering model - Google Patents

Abstract

Disclosed are a route recommending service method and a device for an electric vehicle based on a hybrid orienteering model. The route recommending service method for the electric vehicle includes a step of generating multiple chromosomes corresponding to multiple visitation schedules including multiple visitation spots -the visitation spots including at least one selection spot and at least one recommendation spot- by using an encoding scheme based on at least the selection spot and the recommendation spot; a step of extracting an initial population based on the chromosomes; and a step of extracting an optimized visitation schedule by mating and selecting the chromosomes including the initial population by using genetic algorithm based on the fitness of the chromosomes included in the initial population.

Description

TECHNICAL FIELD [0001] The present invention relates to a route referencing service method and apparatus for an electric vehicle based on a hybrid orientation model,

The following embodiments relate to a method and apparatus for route recommendation service for an electric vehicle based on a hybrid orientation model.

Electric vehicles can not only reduce greenhouse gases but also improve energy efficiency, and are expected to gradually replace existing cars. However, the electric vehicle not only requires a long charging time, but also the traveling distance is relatively shorter than that of the conventional vehicle, and the electric vehicle can be charged only in a limited place. Therefore, when the electric vehicle needs to visit a plurality of visit points, the total travel distance of the electric vehicle and the waiting time required for charging the electric vehicle may be changed according to the visiting order of the electric vehicle. This means that although the same visit point is moved, the resources required for the visit may be different depending on the visit order. Accordingly, in order to efficiently operate an electric vehicle, researches for providing an optimal visit schedule to users are being continued.

Embodiments of the present invention can provide a route recommendation method and apparatus for an electric vehicle capable of providing an optimal visit schedule to a user using an encoding scheme and a genetic algorithm based on a selection point and a recommendable point.

A path recommendation method for an electric vehicle according to an embodiment of the present invention includes the steps of: generating a plurality of visit points using the encoding scheme based on at least one selection point and at least one recommendable point, Generating a plurality of chromosomes corresponding to a plurality of visited schedules including the at least one selection point and the at least one recommendation point; Extracting an initial population based on the plurality of chromosomes; And extracting an optimal visit schedule by selecting and crossing a plurality of chromosomes included in the initial population using a genetic algorithm based on a fitness of the plurality of chromosomes included in the initial population, .

Wherein generating the plurality of chromosomes comprises mapping the at least one selection point and the at least one recommendable point as integers; Generating a plurality of first random sequences based on the mapped at least one selection point; Generating a plurality of second random sequences by classifying the mapped at least one recommendable point as an omission point and a recommendation point based on an omission probability; And generating the plurality of chromosomes using the plurality of first random sequences and the plurality of second random sequences.

Wherein the extracting of the initial population comprises: setting the plurality of chromosomes as the initial population; Randomly arranging the order of the genes of each of the plurality of chromosomes to generate a plurality of new chromosomes; Determining whether the plurality of chromosomes included in the initial population and the plurality of new chromosomes are identical; And including the plurality of new chromosomes in the initial population if the plurality of chromosomes included in the initial population and the plurality of new chromosomes are not identical as a result of the determination.

Wherein the step of determining whether the plurality of chromosomes included in the initial population and the plurality of new chromosomes are identical comprises determining a plurality of chromosomes included in the initial population and a plurality of chromosomes corresponding to the omission point Removing the gene; And comparing the plurality of genes included in the initial population and the plurality of new genes from which the genes corresponding to the omitted points have been removed.

Wherein the extracting of the optimal visit schedule comprises: evaluating the fitness of the plurality of chromosomes included in the initial population; Selecting parents among the plurality of chromosomes included in the initial population according to the fitness; And crossing the selected parents to obtain a child.

The step of obtaining a child by crossing the selected parents includes: determining whether a gene corresponding to a missing visit point exists in the child when the child includes a gene corresponding to a duplicated visit point; And replacing the gene corresponding to the duplicated visit point with a gene corresponding to the missing visit point.

The step of obtaining a child by crossing the selected parent includes: identifying a new recommendable point; And replacing the gene corresponding to the duplicated visit point with a gene corresponding to the new recommendable point.

A path recommendation method for an electric vehicle according to an embodiment of the present invention includes generating a mutant chromosome, which is a chromosome different from a plurality of chromosomes existing in the initial population using a mutation; And setting the mutant chromosome to the child.

The fitness of the child may be improved as compared with the parents.

The step of selecting parents among the plurality of chromosomes may use a Roulette Wheel Selection method.

Wherein evaluating the fitness of the plurality of chromosomes contained in the initial population comprises: identifying a plurality of visit schedules corresponding to the plurality of chromosomes; Calculating a sum of waiting times required for charging the electric vehicle at the plurality of visit points included in each of the plurality of visiting schedules; And evaluating the fitness of the plurality of chromosomes based on the sum of the latencies.

The step of calculating the sum of the waiting times may use the available battery remaining amount, the waiting time at each of the visit points, and the battery remaining amount at the time when the electric vehicle moves.

The initial population may include a predetermined number of the plurality of chromosomes.

An apparatus for recommending a route for an electric vehicle according to an embodiment of the present invention includes a plurality of visit points using an encoding scheme based on at least one select point and at least one recommendable point, A chromosome generator for generating a plurality of chromosomes corresponding to a plurality of visit schedules including a selection point and the at least one recommendation point; An initial population extracting unit for extracting an initial population based on the plurality of chromosomes; And an optimal visit schedule extracting unit for extracting optimal visit schedules by selecting and crossing a plurality of chromosomes included in the initial population using a genetic algorithm based on the fitness of the plurality of chromosomes included in the initial population, .

Wherein the chromosome generating unit comprises: an integer mapping unit for mapping the at least one selection point and the at least one recommendable point as integers; A first random sequence generator for generating a plurality of first random sequences based on the mapped at least one selection point; A second random sequence generator for generating a plurality of second random sequences by classifying the mapped at least one recommendable point as a skipped point and a recommended point based on the skipped probability; And a generation unit that generates the plurality of chromosomes using the plurality of first random sequences and the plurality of second random sequences.

Wherein the initial population extraction unit comprises an initial population setting unit that sets the plurality of chromosomes as the initial population; A new chromosome generating unit for generating a plurality of new chromosomes by randomly arranging the order of genes of the plurality of chromosomes; A determination unit for determining whether the plurality of chromosomes included in the initial population and the plurality of new chromosomes are identical; And an initial population inclusion unit that includes the plurality of new chromosomes in the initial population if the plurality of chromosomes included in the initial population and the plurality of new chromosomes are not identical as a result of the determination.

The optimal visit schedule extracting unit may include a fitness evaluating unit for evaluating the fitness of the plurality of chromosomes included in the initial population; A parent selecting unit for selecting parents among the plurality of chromosomes included in the initial population according to the fitness; And a child acquiring unit for acquiring a child by crossing the selected parents.

Wherein the child acquiring unit includes a missing gene determining unit that determines whether a gene corresponding to a missing visit point exists in the child when the child includes a gene corresponding to a duplicated visit point; And a replacement unit for replacing a gene corresponding to the duplicated visit point with a gene corresponding to the missing visit point.

Wherein the child acquiring unit includes a new and old recommendable point identifying unit that identifies a new recommendable point; And a replacement unit for replacing a gene corresponding to the duplicated visit point with a gene corresponding to the new recommendable point.

Wherein the fitness evaluation unit comprises: a plurality of visit schedule identification units for identifying a plurality of visit schedules corresponding to the plurality of chromosomes; A waiting time calculation unit for calculating a sum of waiting times required for charging the electric vehicle at the plurality of visit points included in each of the plurality of visiting schedules; And an evaluation unit for evaluating the fitness of the plurality of chromosomes based on the sum of the waiting times.

Embodiments of the present invention can provide a route recommendation method and apparatus for an electric vehicle capable of providing an optimal visit schedule to a user using an encoding scheme and a genetic algorithm based on a selection point and a recommendable point.

1 is a view for explaining a visiting path of an electric vehicle according to an embodiment.
2 is a flowchart illustrating a route recommendation method for an electric vehicle according to an exemplary embodiment of the present invention.
3 is a diagram for explaining an encoding scheme and a plurality of chromosomes according to an embodiment.
4 is a block diagram illustrating a route recommendation apparatus for an electric vehicle according to an embodiment.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to or limited by the embodiments. In addition, the same reference numerals shown in the drawings denote the same members.

1 is a view for explaining a visiting path of an electric vehicle according to an embodiment.

Referring to FIG. 1, when visiting a plurality of points using an electric vehicle, a user may select points to visit and determine a visit order or a visit schedule. At this time, the waiting time may be an important factor for determining the visit time. Here, the waiting time may mean the difference between the staying time at the visit point and the required charging time of the electric vehicle. And, the residence time may mean the time required for the user of the electric vehicle to perform an arbitrary activity at the visit point. At this time, in one embodiment, the residence times at each of the visited points are not only realistically identifiable but also statistically identifiable. For example, the statistical residence time can be set to an average value of residence times. And, in order to calculate the waiting time, the availability of the charging facility at each visit point can be considered. For example, if the point of visit includes a charging facility, the electric vehicle may be charged during the time the user is staying. At this time, the chargeable amount can be proportional to the residence time at the charging point. Then, the average residence time can be known in advance. At this time, if the remaining battery power of the electric vehicle is sufficient to arrive at the next point, the waiting time may be unnecessary. However, if the remaining battery power of the electric vehicle is not sufficient to reach the next point, a waiting time may occur because the battery must be additionally charged to move to the next place. In the example of Fig. 1 (a), points A, B, and I are selected points that are visited by the user among a plurality of points. If the selection points are point A and point B, the electric vehicle can move from point A to point B, and if point I is added as a selection point, the electric car can move from point A to point B via point I have. At this point, points A and B may include a charging facility, while point I may not include a charging facility. Thus, at point I, the electric vehicle can not be charged. In this case, the electric vehicle may have to be charged at some point along the path. At this time, since the user has to wait for the car to be charged without any activity, the user's time may be wasted. On the other hand, when visiting an additional place with a charging facility, but not at a point of selection, the electric vehicle can perform charging during the residence time, thereby wasting the user's time.

Specifically, the recommendable point may not be the point selected by the user as the point of visit, but may include the charging facility and may be a point at which the user can stay. For example, a recommendable point may include a point, a restaurant, a rest area, and the like, which are not a selection point. In the example of Fig. 1 (b), points A, B and C may mean a selection point. The points R ₁ and R ₂ may refer to recommendable points. For example, point R ₁ and point R ₂ may be extracted from a POI (Point of Interests) database. In this case, in one embodiment, the path of the electric vehicle may be considered as the battery remaining amount, the estimated arrival time at each point, and the amount of battery necessary to move to the next point.

에서

로 변경될 수 있다. 구체적으로, 공간 탐색 복잡도(Space search complexity)를 예측하기 위하여, n 개의 선택 지점들과 m개의 추천 지점들이 존재함을 가정할 수 있다. 이 경우,

개의 경로 추천들이 존재할 수 있고, 각 경로 추천들은

의 카디널리티(Cardinality)를 갖을 수 있다. 이에 따라, 전기 자동차는

개의 방문 순서들을 갖을 수 있다. 이 때, 선택 장소들은 반드시 방문 스케쥴에 포함될 수 있으나, 추천 가능 지점은 방문 스케쥴에 포함되지 못할 수도 있다. 그리고, 최종 방문 스케쥴이 추출되었으나, 전기 자동차의 사용자가 이에 만족하지 못하는 경우, 사용자는 복수의 방문 지점들에서 선택 지점을 다시 결정할 수 있다. 이에 따라, 최적 방문 스케쥴을 결정하는 것은 대기 시간의 감소 및 사용자의 만족을 위하여, 사용자가 선택한 장소를 위한 TSP(Traveling Salesman Problem) 해결자(solver) 및 추천 가능 지점을 위한 오리엔티어링 문제 해결자(Orienteering problem solver)가 조합된 것일 수 있다.
When the recommendable point R ₂ is added to the visit point, as shown in Fig. 1 (c), the visit path is

in

. ≪ / RTI > Specifically, in order to predict the space search complexity, it is assumed that there are n selection points and m recommendation points. in this case,

There may be < RTI ID = 0.0 > a < / RTI >

Of cardinality. As a result,

You can have a sequence of visits. At this time, the selected places may be included in the visit schedule, but the recommendable points may not be included in the visit schedule. If the final visit schedule has been extracted but the user of the electric vehicle is not satisfied with this, the user can re-select the selection point at a plurality of visit points. Accordingly, for determining the optimal visit schedule, a TSP (Traveling Salesman Problem) solver for a place selected by the user and an Orienteering problem solver for recommendable points problem solver may be combined.

2 is a flowchart illustrating a route recommendation method for an electric vehicle according to an exemplary embodiment of the present invention.

Referring to FIG. 2, a route recommendation method for an electric vehicle according to an embodiment (hereinafter referred to as a route recommendation method for an electric vehicle) is based on an encoding scheme ) To generate a plurality of chromosomes corresponding to a plurality of visit schedules including a plurality of visited points (210). Here, the visiting points may include at least one selection point and at least one recommendation point. Here, the selection point may mean a point selected by the user as a visit point among a plurality of points. The recommendable point may be a point at which the user can stay while including the charging facility although the user has not selected the point of visit. In this case, the recommendable points can be classified as recommendation points recommended for the electric vehicle and recommendation points for the electric vehicle, and omitted points not recommended as the visit points.

Specifically, a genetic algorithm based on the principle of natural selection and genetics can be used to extract an optimal visit schedule, a path recommendation method for an electric vehicle. To implement this genetic algorithm, a path recommendation method for an electric vehicle can encode each visit schedule into a chromosome using an encoding scheme. At this time, the gene may be an integer-valued vector.

More specifically, step 210 may map the at least one selection point and the at least one recommendable point as integers. For example, if there are five select points and three recommendable points, then step 210 may map each select point to the integers 0-4 and map each recommendable point to 6-8 .

And, step 210 may generate a plurality of first random sequences based on the mapped at least one selection point. For example, if there are five selection points, step 210 may be performed by randomly arranging the order of the selection points (0, 1, 2, 3, 4), such as (3, 2, 4, 1, 0) 5 < th > random sequence. The step 210 may generate a plurality of second random sequences by classifying at least one recommendable point mapped on the basis of an omission probability into an omission point and a recommendation point. Here, the skip probability may mean the ratio of the skipped point and the recommended point among the recommendable points. In one embodiment, the omission probability can be predetermined, and the omission point can be mapped to -1. For example, if there are three recommendable points and the elliptical probability is 1/3, step 210 can classify one of the three recommendable points as an omission point and two points as a recommendation point have. Accordingly, step 210 may generate a second random sequence such as (6, -1, 8), (6, 7, -1)

Then, step 210 may generate a plurality of chromosomes using a plurality of first random sequences and a plurality of second random sequences. Specifically, step 210 may combine a plurality of first random sequences and a plurality of second random sequences to represent chromosomes in the form of integer-value vectors. In the example described above, step 210 may be performed by combining the first random sequence and the second random sequence so that (0,1,2,3,4,6, -1,8), (3,2,4,1 , 0, 6, -1, 8), and the like. In this case, the length of a plurality of chromosomes may be eight.

And, in the TSP, if there are no constraints, a number of invalid chromosomes can be included in the population. Here, an ineffective chromosome may include a chromosome that does not include a gene corresponding to a selection point, a chromosome that includes a gene corresponding to the same visit point, and the like. Accordingly, step 210 may include constraints that can generate only valid chromosomes. Although this constraint narrows the search range and may adhere to the local minimum, it can quickly converge the genetic loop, which will be described later. On the other hand, in the absence of such constraints, path recommendation methods for electric vehicles can take a very long time to process ineffective chromosomes and fail to extract any quality solution.

In addition, the path recommendation method for an electric vehicle can extract an initial population based on a plurality of chromosomes (220). Here, the initial population may include a predetermined number of plural chromosomes. In addition, the path recommendation method for the electric vehicle can extract an optimal visit schedule by selecting and crossing a plurality of chromosomes included in the initial population.

Specifically, step 220 may set a plurality of chromosomes as the initial population. Then, in step 220, a plurality of new chromosomes can be generated by randomly arranging the order of the genes of the plurality of chromosomes. For example, when there are 5 selection points and 3 recommendable points, one of the genes in the initial population is (0, 1, 2, 3, 4, 6, -1, 8 ). In this case, the step 220 is performed on the basis of the chromosomes 0 (0, 0, 1, 0, 4, 3, 6, 1, 8) , 1, 2, 3, 4, 6, -1, 8) may be randomly arranged to generate a plurality of new chromosomes.

Step 220 may determine whether a plurality of chromosomes included in the initial population and a plurality of new chromosomes are identical. More specifically, step 220 may remove a plurality of chromosomes included in the initial population and a gene corresponding to the omitted point for the plurality of new chromosomes. The plurality of chromosomes included in the initial population, from which the gene corresponding to the omitted point is removed, and the plurality of new chromosomes can be compared. This may be because the omission point is excluded from the actual visit schedule. For example, (A, B, -1, C) and (A, -1, B, C) can all refer to schedules of electric cars visited in the order of A → B → C. Thus, (A, B, -1, C) and (A, -1, B, C) can mean the same schedule.

As a result of the determination, if a plurality of chromosomes included in the initial population and a plurality of new chromosomes are not identical, the step 220 may include a plurality of new chromosomes in the initial population. Accordingly, a plurality of new chromosomes identical to a plurality of chromosomes contained in the initial population can be removed. And, in one embodiment, step 220 may repeatedly generate a new chromosome and compare it to a plurality of chromosomes contained in the initial population, until the initial population includes a predetermined number of chromosomes.

In addition, a path recommendation method for an electric vehicle includes selecting and crossing a plurality of chromosomes included in the initial population using a genetic algorithm based on the fitness of the plurality of chromosomes included in the initial population, (230). First, step 230 may evaluate the fitness of a plurality of chromosomes included in the initial population. Here, the fitness may be based on the waiting time for charging the electric vehicle. Accordingly, a chromosome corresponding to a visit schedule with a small waiting time can be evaluated as having a high fitness. At this time, one embodiment may use a fitness function to evaluate the fitness of a plurality of chromosomes.

Specifically, step 230 may identify a plurality of visited schedules corresponding to a plurality of chromosomes. Then, the sum of the waiting times at the plurality of visit points included in each of the plurality of visit schedules can be calculated. To this end, one embodiment may assume that the battery is buffered when the electric vehicle starts its first movement, and the mean residence time of the electric vehicle is known. The waiting time at each of the visited points can be calculated using the following equation (1).

[Equation 1]

는 방문 지점,

는 현재 사용 가능한 배터리 잔량,

는 전기 자동차가 방문 지점

에 도착했을 때의 거리 지수(distance credit),

는 최대 배터리 용량,

는 방문 지점

에서의 체류 시간을 의미할 수 있다. 그리고,

는 방문 지점

에서의 대기 시간,

는 이전 방문 지점에서의 대기 시간,

는

와

사이의 로드 네트워크 거리(Road network distance)를 의미할 수 있다. 그리고,

은 전기 자동차가 이동할 시점의 배터리 잔량을 의미할 수 있다. 이 때, 방문 지점

에서의 체류 시간이 존재하지 않을 경우,

는 0 일 수 있고, 이에 따라, 방문 지점

에서의 대기 시간

는 증가할 수 있다.here,

A visit point,

Currently available battery level,

The electric vehicle

Distance credit,

The maximum battery capacity,

The point of visit

Can be regarded as the residence time at the point of time. And,

The point of visit

The waiting time at,

Is the waiting time at the previous visit point,

The

Wow

And may be referred to as a " road network distance ". And,

May mean the remaining battery level at the time when the electric vehicle moves. At this time,

If there is no residence time in the room,

May be zero, and accordingly,

Waiting time at

Can be increased.

The sum of the waiting times at the plurality of visited points can be expressed by the following equation (2).

&Quot; (2) "

는 복수의 방문 지점들에서의 대기 시간의 합을 의미할 수 있다. 이에 따라, 단계 (230)는 복수의 방문 스케쥴들 각각에 포함된 복수의 방문 지점들에서의 대기 시간의 합을 계산할 수 있다.here,

May mean the sum of the waiting times at a plurality of visited points. Thus, step 230 may calculate the sum of the waiting times at the plurality of visited points included in each of the plurality of visited schedules.

Then, step 230 may evaluate the fitness of the plurality of chromosomes based on the sum of the waiting times. For example, chromosomes with a small sum of latency can be evaluated as having high fitness, and chromosomes with a large sum of latency can be evaluated as having low fitness.

Then, step 230 may select the parent among the plurality of chromosomes included in the initial population according to the fitness. In particular, step 230 may extract the optimal visit schedule using a genetic loop. Here, the genetic loop can mean selection, mating, and mutation by fitness. And, by repetition of the genetic loop, chromosomes more and more good in fitness can be generated from the initial population. This may be because the chromosomes with the highest fitness in the initial population are selected as parents, the selected parents are mated to obtain the offspring, and only the chromosomes with the highest fitness among the newly generated children survive in the population. And, one embodiment may use a Roulette Wheel Selection method to select parents among a plurality of chromosomes. Using the roulette wheel selection method as described above, it is possible to increase the probability of acquiring a child having a higher fitness in crossing

Then, step 230 may obtain the child by crossing the selected parents. Here, crossing may be a process for obtaining a child with a better fit. Specifically, mating can select a pair of intersections randomly on the chromosomes of the parents, and proceed with the process of changing the partial sequence of the parents' chromosomes based on the intersection. At this time, in one embodiment, the child may include a gene corresponding to a duplicated visit point. This may mean visiting a particular visit point more than once in the visit schedule, and such visit schedule may not be valid. For this purpose, step 230 can determine whether there is a gene corresponding to the missing visit point in the child, and if there is a gene corresponding to the missing visit point, It can be replaced with a gene corresponding to the visiting point. Accordingly, an ineffective chromosome can be an effective chromosome without missing a visit point. And, in one embodiment, the duplicate visited points and the missing visited points may not be matched. In this case, step 230 may identify a new recommendable point. By replacing a gene corresponding to a duplicated visit point with a gene corresponding to a new recommendable point, an invalid chromosome can be an effective chromosome to which a new recommendable point is added.

In step 230, a mutation chromosome, which is a chromosome different from a plurality of chromosomes existing in the initial population, may be generated using a mutation. Specifically, a mutation is a technique for randomly exchanging genes in a chromosome, which can prevent the path recommendation method for an electric vehicle from sticking to a local minimum. A common search technique can generate mutations at a predetermined rate. However, the path recommendation method for an electric vehicle can determine whether a new chromosome is the same as an existing chromosome, and, in the same case, can remove a new chromosome. In addition, an ineffective chromosome can also be removed. Accordingly, step 230 can generate mutant chromosomes that are chromosomes different from a plurality of chromosomes. At this time, the mutant chromosome may be an effective chromosome. In this case, in the generation of the mutation chromosome, the gene corresponding to the selection point and the gene corresponding to the recommended point may be different from each other. Then, the step 230 may attempt to mutate until a chromosome completely different from an existing chromosome is extracted. Then, step 230 can set the mutation chromosome as a child and repeat the genetic loop until the optimal visit schedule is extracted.

Therefore, the route recommendation method for the electric vehicle can repeat the genetic loop until the chromosome having the highest fitness is extracted, and the chromosome with the highest fitness can be set as the optimum chromosome. The visit schedule corresponding to the optimal chromosome can be set as the optimal visit schedule.

3 is a diagram for explaining an encoding scheme and a plurality of chromosomes according to an embodiment.

Referring to FIG. 3, FIGS. 3A and 3B may illustrate the encoding scheme and the generation of a plurality of chromosomes. FIG. Specifically, in the example of FIG. 3 (a), the number of selection points may be eight, and the number of recommendable points may be five. Accordingly, the length of the integer-value vector generated later can be 13. The path recommendation method for an electric vehicle can generate a plurality of first random sequences by mapping selected points to integers from 0 to 7, map recommendable points to integers from 9 to 13, A plurality of second random sequences may be generated by classifying the recommendable points as the omitted points and the recommended points. In this case, the omitted points may be mapped to -1, and the omitted points may be indicated by '-' in FIG. 3 (a) and FIG. 3 (b). The path recommendation method for the electric vehicle may be represented by a chromosome of integer-value vector type as shown in FIG. 3 (b) by combining a plurality of first random sequences and a plurality of second random sequences.

In the example of FIG. 3 (b), each row may mean a respective chromosome. For example, in the case of a chromosome with an integer-value vector of (4, 3, 11, 1, 6, -, 0, 7, 9, 5, -, 2, 13) 11 (recommendation point 3) → ... → point 2 → point 13 (recommendation point 5). Such a visit schedule may be a valid schedule where the visit points do not overlap. A path recommendation method for an electric vehicle can generate a plurality of new chromosomes by randomly arranging the order of the genes of the plurality of chromosomes. If the new chromosome overlaps with the existing chromosome, the new chromosome can be removed. Accordingly, the path recommendation method for the electric vehicle can continue the generation of the encoding scheme and the plurality of chromosomes shown in Figs. 3 (a) and 3 (b) until the generation of the initial population is completed.

4 is a block diagram illustrating a route recommendation apparatus for an electric vehicle according to an embodiment.

The chromosome generating unit 410 can generate a plurality of chromosomes corresponding to a plurality of visit schedules including a plurality of visit points using the encoding scheme based on the at least one selection point and the at least one recommendable point .

In addition, the initial population extraction unit 420 may extract an initial population based on a plurality of chromosomes.

Also, the optimal visit schedule extracting unit may extract an optimal visit schedule by selecting and crossing a plurality of chromosomes included in the initial population using a genetic algorithm based on the fitness of a plurality of chromosomes contained in the initial population.

The route recommendation apparatus for an electric vehicle shown in FIG. 4 can be applied to the route guidance system shown in FIG. 1 through FIG. 3 as it is, so that detailed description will be omitted.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (20)

A plurality of visit points using an encoding scheme based on at least one select point and at least one recommendable point, the visit points including the at least one select point and the at least one recommendation point - generating a plurality of chromosomes corresponding to a plurality of visited schedules comprising;
Extracting an initial population based on the plurality of chromosomes; And
Extracting an optimal visit schedule by selecting and crossing a plurality of chromosomes contained in the initial population using a genetic algorithm based on a fitness of the plurality of chromosomes included in the initial population;
And a route recommendation method for recommending a route for an electric vehicle.
The method according to claim 1,
The step of generating the plurality of chromosomes
Mapping the at least one selection point and the at least one recommendable point as integers;
Generating a plurality of first random sequences based on the mapped at least one selection point;
Generating a plurality of second random sequences by classifying the mapped at least one recommendable point as an omission point and a recommendation point based on an omission probability; And
Generating the plurality of chromosomes using the plurality of first random sequences and the plurality of second random sequences
And a route recommendation method for recommending a route for an electric vehicle.
3. The method of claim 2,
The step of extracting the initial population
Setting the plurality of chromosomes as the initial population;
Randomly arranging the order of the genes of each of the plurality of chromosomes to generate a plurality of new chromosomes;
Determining whether the plurality of chromosomes included in the initial population and the plurality of new chromosomes are identical; And
If the plurality of chromosomes included in the initial population and the plurality of new chromosomes are not the same, the step of including the plurality of new chromosomes in the initial population
And a route recommendation method for recommending a route for an electric vehicle.
The method of claim 3,
Wherein the step of determining whether the plurality of chromosomes included in the initial population and the plurality of new chromosomes are identical
Removing a gene corresponding to the omission point with respect to a plurality of chromosomes included in the initial population and the plurality of new chromosomes; And
Comparing the plurality of genes included in the initial population and the plurality of new genes from which the genes corresponding to the omitted points have been removed,
And a route recommendation method for recommending a route for an electric vehicle.
The method according to claim 1,
The step of extracting the optimal visit schedule
Evaluating the fitness of the plurality of chromosomes contained in the initial population;
Selecting parents among the plurality of chromosomes included in the initial population according to the fitness; And
The step of obtaining a child by crossing the selected parents
And a route recommendation method for recommending a route for an electric vehicle.
6. The method of claim 5,
The step of crossing the selected parents to obtain a child
If the child includes a gene corresponding to a duplicated visit point,
Determining whether the child has a gene corresponding to a missing visit point; And
Replacing the gene corresponding to the duplicated visit point with a gene corresponding to the missing visit point
And a route recommendation method for recommending a route for an electric vehicle.
The method according to claim 6,
The step of obtaining a child by crossing the selected parent
Identifying a new recommendable point; And
Replacing a gene corresponding to the duplicated visit point with a gene corresponding to the new recommendable point
And a route recommendation method for recommending a route for an electric vehicle.
6. The method of claim 5,
Generating a mutant chromosome that is a chromosome different from a plurality of chromosomes existing in the initial population using a mutation; And
Setting the mutant chromosome to the child
And a route recommendation method for recommending a route for an electric vehicle.
6. The method of claim 5,
The child
And a path recommendation apparatus for recommending a path for an electric vehicle of the path recommendation apparatus improved in fitness than the parents.
9. The method of claim 8,
The step of selecting the parents among the plurality of chromosomes
A route recommendation method for an electric vehicle of a route recommendation apparatus using a roulette wheel selection method.
6. The method of claim 5,
Wherein evaluating the fitness of the plurality of chromosomes contained in the initial population comprises
Identifying a plurality of visited schedules corresponding to the plurality of chromosomes;
Calculating a sum of waiting times required for charging the electric vehicle at the plurality of visit points included in each of the plurality of visiting schedules; And
Evaluating the fitness of the plurality of chromosomes based on the sum of the waiting times
And a route recommendation method for recommending a route for an electric vehicle.
12. The method of claim 11,
The step of calculating the sum of the waiting times
A route recommendation apparatus for recommending a route for an electric vehicle, the route recommendation apparatus using the available battery remaining amount, the waiting time at each of the visit points, and the battery remaining amount at the time when the electric vehicle moves.
The method according to claim 1,
The initial population
A path recommendation method for an electric vehicle of a path recommendation apparatus including a predetermined number of the plurality of chromosomes.
Using a coding scheme based on at least one selection point and at least one recommendable point, the plurality of points of visit including the at least one selection point and the at least one recommendation point A chromosome generating unit for generating a plurality of chromosomes corresponding to a plurality of visited schedules;
An initial population extracting unit for extracting an initial population based on the plurality of chromosomes; And
An optimal visit schedule extracting unit for extracting optimal visit schedules by selecting and crossing a plurality of chromosomes included in the initial population using a genetic algorithm based on the fitness of the plurality of chromosomes included in the initial population,
And a route recommending device for an electric vehicle.
15. The method of claim 14,
The chromosome generating unit
An integer mapping unit for mapping the at least one selection point and the at least one recommendable point as integers;
A first random sequence generator for generating a plurality of first random sequences based on the mapped at least one selection point;
A second random sequence generator for generating a plurality of second random sequences by classifying the mapped at least one recommendable point as a skipped point and a recommended point based on the skipped probability; And
Generating a plurality of chromosomes using the plurality of first random sequences and the plurality of second random sequences,
And a route recommending device for an electric vehicle.
15. The method of claim 14,
The initial population extraction unit
An initial population setting unit setting the plurality of chromosomes as the initial population;
A new chromosome generating unit for generating a plurality of new chromosomes by randomly arranging the order of genes of the plurality of chromosomes;
A determination unit for determining whether the plurality of chromosomes included in the initial population and the plurality of new chromosomes are identical; And
If it is determined that the plurality of new chromosomes are not identical to the plurality of new chromosomes included in the initial population and that the plurality of new chromosomes are included in the initial population,
And a route recommending device for an electric vehicle.
15. The method of claim 14,
Wherein the extracting unit extracts the optimal visit schedule
A fitness evaluation unit for evaluating the fitness of the plurality of chromosomes included in the initial population;
A parent selecting unit for selecting parents among the plurality of chromosomes included in the initial population according to the fitness; And
A child acquiring unit for acquiring a child by crossing the selected parents,
And a route recommending device for an electric vehicle.
18. The method of claim 17,
The child acquiring unit
If the child includes a gene corresponding to a duplicated visit point,
A missing gene determining unit for determining whether or not a gene corresponding to a missing visit point exists in the child; And
And replacing the gene corresponding to the duplicated visit point with a gene corresponding to the missing visit point
And a route recommending device for an electric vehicle.
19. The method of claim 18,
The child acquiring unit
A new and old recommendable point identification unit for identifying a new recommendable point; And
And a replacement unit for replacing a gene corresponding to the duplicated visit point with a gene corresponding to the new recommendable point,
And a route recommending device for an electric vehicle.
18. The method of claim 17,
The fitness evaluation unit
A plurality of visit schedule identifying units for identifying a plurality of visit schedules corresponding to the plurality of chromosomes;
A waiting time calculation unit for calculating a sum of waiting times required for charging the electric vehicle at the plurality of visit points included in each of the plurality of visiting schedules; And
And evaluating the fitness of the plurality of chromosomes based on the sum of the waiting times.
And a route recommending device for an electric vehicle.

Images