KR102478210B1 - Charging device that optimizes the operation of the charging system in the charging area - Google Patents

Abstract

A charging device of the present invention may comprise: a plurality of chargers in which a queue is formed in which an electric car of a customer waits before charging, and the electric car can be charged; a server for receiving data on a distribution of arrival intervals of electric vehicles entering the queue and data on a distribution of charging times in which the electric vehicles are charged at the charger; and a control unit for setting an objective function, which is the total cost invested in design or operation, using the data received by the server. The charging device of the present invention can consider a customer patience time following an independently and identically arbitrary distribution when setting the objective function. The customer patience time may mean a queue churn rate of customers who give up charging without being provided with the charging of the electric vehicles of the customers when the waiting time of the electric vehicle of the customer in the queue exceeds the customer patience time. The control unit may calculate the number of chargers for an electric vehicle that optimizes the objective function.

Description

Charging device that optimizes the operation of the charging system in the charging area}

The present invention relates to a charging device that optimizes the operation of a charging system in a charging area for cost optimization associated with charging an electric vehicle.

Unlike the charging of vehicles using fossil fuels, charging of electric vehicles may take longer. Therefore, the charging device for charging the electric vehicle is modeled as a stochastic process including the arrival process of the electric vehicle charging device to the queue, the service process, the queue state, or the vehicle state, etc. to achieve optimal charging for both revenue generation and customer satisfaction. Can be selected to configure the station.

A charging device for optimizing the operation of a charging system in a charging zone of the present invention relates to optimizing an objective function, which is the total cost including the length of an electric vehicle queue, waiting time, or charging device.

The charging device of the present invention forms a queue in which a customer's electric vehicle waits for charging, a plurality of chargers capable of charging electric vehicles, data on the distribution of arrival intervals of electric vehicles entering the queue, and charging in which the electric vehicles are charged at the charger. It may include a server that receives data on time distribution, and a control unit that sets an objective function, which is the total cost invested in design or operation, using the data received by the server.

The charging device of the present invention may consider the customer patience time following an arbitrary distribution that is independently and identically when setting the objective function, and the customer patience time is the customer patience time waiting time for the customer's electric vehicle in the queue. If it exceeds, the customer's electric vehicle may indicate a queue dropout rate of customers who give up charging without being provided with charging, and the control unit may calculate the number of electric vehicle chargers that optimize the objective function.

The control unit of the present invention may receive an input of the arrival interval distribution of vehicles entering the queue, the charging time distribution, or the number of chargers, and the control unit may receive input of the waiting time of the electric vehicle waiting in the queue or the electric vehicle waiting in the queue. It is possible to calculate the queue length, which is the number of

The arrival interval distribution of the present invention may follow an arbitrary distribution that is independently and identically, and the charging time distribution may follow an independent phase-type distribution.

The objective function, which is the total cost of the present invention, may be the social input cost. The objective function may include at least one of charging station investment cost, labor cost, charging station maintenance cost, charger failure cost, customer purchase cost, rapid charging cost, and customer waiting time. Each cost of the objective function can be calculated by the number of chargers (n).

1 is a configuration diagram of a charging device of the present invention.
2 is an operation explanatory diagram of the charging device of the present invention.

In the representation of an A/B/n/K queue system, A may be a distribution of arrival time intervals, B may be a distribution of charging times, n may be the number of chargers, or K may be an acceptable queue length. May be omitted if the length of the acceptable queue is infinite.

For example, M/G/1 queue can mean that one charger and the arrival time interval follow an exponential distribution, and the charging time distribution follows a random distribution, and G/G/1 queue means one charger and arrival time Intervals and filling times can be meant to follow arbitrary distributions.

A stochastic process can be said to be a group of random variables that have a specific value at a specific point in time and are sequentially observed. If a stochastic process at any point in time is affected only by the stochastic process at the immediately preceding point in time, a discrete-time stochastic process with Markov properties is called a Markov chain, and a continuous-time stochastic process is called a Markov process. can be said When the time intervals in which events occur are independent and follow the same exponential distribution, such a Markov process can be called a Poisson process.

That is, the Markov stochastic process may be a stochastic process in which the past and the future are independent of each other as a condition for the present, and the Markov stochastic process may be a 'non-remembering probability' process. If we are trying to infer the future from a Markov probability process, only the present values are useful, and the past values provide no additional information.

An embodiment of the charging device of the present invention may be a G/Ph/n + G queuing system.

The first letter, G, may be an arrival time interval distribution and may mean following an arbitrary distribution that is independently and identically identical. The charging time distribution may follow an independent uniform phase-type distribution. The topological type distribution can be a probability distribution consisting of a mixture of convolutional or exponential distributions. A phase-type distribution may occur in a system in which one or more 'interrelated Poisson processes occur sequentially or stepwise, and the order in which each step occurs may itself be a stochastic process. Also, + G can be the customer's patience time and can follow an arbitrary distribution with independent equality.

A G/Ph/n + G queuing system can continue to treat changes in customer encounter density over time as an exogenous process. n identical chargers can be placed in parallel, and the customer arrival process can be an update process. Customers can begin service immediately if an idle charger is available upon arrival. Otherwise, customers can wait in queues with infinite waiting rooms holding first-in-first-out queues.

The charge time may form a random variable following a series of independently and identically distributed according to the phase type distribution. Once the charger has finished servicing the customer, the charger can fetch the leading customer from the queue. The charger can be idle if the queue is empty.

Each customer can be patient, and patience times can follow a normal distribution. If the amount of time a customer waits in the queue exceeds the patience time, the customer may forgo the charge without being offered a charge. That is, you can leave the queue.

The operation of the charging device for optimizing the operation of the charging system in the charging zone of the present invention will be described.

The operation or charging method of the charging device that optimizes the operation of the charging system in the charging area is a data collection step (S10), a raw data acquisition step (S100), an objective function (C) setting step (S200), or an objective function (C) An optimization step (S300) may be included.

The data collected in the data collection step (S10) includes information on whether the electric vehicle is charging or idle in a plurality of chargers installed in the charging device, and the charging time required for the electric vehicle 100 to charge when charging. Information about the vehicle, or information about the arrival interval of a vehicle entering the queue from outside may be included.

The objective function (C) may be the social input cost (C). The objective function (C) includes the charging station investment cost (C _fc ), labor cost (C _ps ), charging station maintenance cost (C _dm ), charger failure cost (C _bm ), customer purchase cost (C _be ), and rapid charging cost (C _dw ), and at least one of the customer's waiting time (C _aw ).

In this case, the data in the data collection step (S10) includes charging station investment cost (C _fc ), labor cost (C _ps ), charging station maintenance cost (C _dm ), charger failure cost (C _bm ), customer purchase cost (C _be ), At least one of a fast charging cost (C _dw ) and a customer's waiting time (C _aw ) may be included.

The raw data acquisition step (S100) may include an arrival interval or charging time distribution extraction step (S110), a charger number input step (S120), or an arrival rate (λ) or charging rate (μ) calculation step (S130).

An objective function (C) configured for optimization of operation or deployment of charging devices may include waiting time in a queue of a customer's electric vehicle. Optimization of the objective function (C) may include minimizing the waiting time (Wq). The waiting time (Wq) of the queue may be obtained from at least one of a queue length (Lq), an arrival rate (λ), a filling rate (μ), a utilization rate (ρ), or an idle rate (1-ρ).

Therefore, the step of extracting the arrival interval or charging time distribution (S110) is a step of extracting a target data distribution from the raw data of the arrival interval or charging time to obtain information on the arrival rate (λ) or the charging rate (μ). can be

The control unit 320 may calculate the arrival rate (λ) or the charging rate (μ) based on the arrival interval distribution or charging time distribution transmitted to the server 300, or information on the current number of chargers placed in the corresponding charging device. there is.

The objective function (C) setting step (S100) may be performed by the controller 320. The objective function (C) setting step (S100) includes the utilization rate (ρ) or idle rate (1-ρ) calculation step (S210), queue length calculation step (S220), waiting time (Caw) calculation step (S230), It may include at least one of a cost calculation step (S240), or a customer's patience time or customer churn rate (a) calculation step.

Consequently, the input parameter of the objective function (C) may be at least one of the arrival rate, the filling rate, the number of chargers, or the customer's patience time, and the output parameter of the objective function (C) may be the length of the queue or the number of customers in the queue. It may be a waiting time (Caw). In this process, the controller 320 may calculate a utilization rate, which is a usage rate of the charger, or an idle day, which is a rate where the charger is not used.

The arrival rate (λ) can be given by the relationship of Equation 1 below.

는 브라운 운동(Brownian motion)을 따를 수 있고, 거의 확실하게 연속적인 샘플경로를 갖는 연속 시간 마코프 프로세스일 수 있다.Here, E ⁿ may be the arrival process of the nth system, λ ⁿ may be the arrival rate of the nth system,

may follow Brownian motion, and may almost certainly be a continuous-time Markov process with a continuous sample path.

The arrival rate (λ) and the filling rate (μ) may have a relationship of Equation 2 below.

K 인 하위확률행렬(sub-stochastic matrix)일 수 있다. 이때, P의 대각성분은 0이라고 가정할 수 있다. The charging factor μ may be an average charging factor and may follow a phase-type distribution. It may be a distribution having a phase in which K is 1 or more, and each phase-type may have p, ν, and P as parameters. P may be a K-dimensional vector having a value of 0 or more, ν may be a K-dimensional vector consisting of positive numbers, and P is K

It may be a sub-stochastic matrix of K. At this time, it can be assumed that the diagonal component of P is 0.

A continuous phase-type random variable having p, ν, and P as parameters may follow Ph(p, ν, P), with an initial distribution p and a continuous-time Markov chain in which the velocity matrix G reaches state K+1 ( It can be defined as the first time until the Markov Chain).

The velocity matrix G can be given by Equation 3 below.

는 크기가 K X K인 행렬일 수 있고, h(=-Fe)는 K 차원의 벡터일 수 있다. here,

may be a matrix of size KXK, and h(=-Fe) may be a K-dimensional vector.

The number of chargers may be given as n.

If the customer's waiting time in the queue exceeds the customer's patience time (F), the customer may leave without receiving service. The customer's patience time (F) may follow an independently and identically probability distribution from the time the customer arrives. We can assume that a customer's patience times follow some general distribution.

The customer's patience time (F) may satisfy Equation 4 below.

The utilization rate (ρ) may be defined as ρ = λ / (n * μ), and n may be the number of chargers. The utilization rate may be the utilization rate of the charger and may be the rate at which the charger is in use. The idle rate may be 1 - ρ, and may be a rate at which the charger is not used. The queue length may be an average of the number of customers in the queue waiting for charging at the charging station. The waiting time Caw may be the waiting time of the customer in the queue 22 .

The queue utilization rate or utilization rate ρ may be the utilization rate for a charger, and may be a percentage that a charger is in use. The utilization factor of the n-th charger may be given by ρ ⁿ = λ ⁿ / (n * μ), and may satisfy Equation 5 below.

는 시간 t에 대한 위상(phase) k 서비스에서 n번째 충전기의 고객 수일 수 있고,

는 충전기 할당 과정(server-allocation process)일 수 있다.

may be the number of customers of the nth charger in phase k service for time t,

may be a charger allocation process (server-allocation process).

는 t 시간일 때, n번째 충전기에서의 고객수일 수 있다.

는 n번째 충전기에서 총 고객 수 과정(total-customer-count process)일 수 있고, 다음의 수학식 6을 만족할 수 있다.

may be the number of customers at the n-th charger at time t.

may be a total-customer-count process in the n-th charger, and may satisfy Equation 6 below.

Here, F may be the customer's patience time or the customer churn rate.

가 시간 t 일때 사용하지 않는 충전기의 개수인 경우, 다음의 수학식 7이 만족할 수 있다.

When is the number of unused chargers at time t, Equation 7 below may be satisfied.

가 시간 t 일때 대기열에서 기다리는 고객의 수인 경우 다음의 수학식 8이 만족할 수 있다.

When is the number of customers waiting in the queue at time t, Equation 8 below can be satisfied.

는 시간 t일 때 가상의 무한대의 인내시간을 가진 고객이 가지는 잠재적인 대기 시간일 수 있다.

may be the potential waiting time of a customer with a virtually infinite patience at time t.

Accordingly, the length or waiting time of the queue may be determined or calculated by the number n of chargers. The objective function (C) to be optimized, that is, the total cost may include the length or waiting time of a queue, and an embodiment of a method of optimizing the objective function (C) is the number of chargers that minimize the objective function (C). It can be the same as finding (n).

The objective function (C) may be set as in Equation 9.

Here, C may be an objective function or a social input cost. Cfc is the charging station investment cost, Cps is the labor cost, Cdm is the charging station maintenance cost, Cbm is the charger failure cost, Cbe is the customer's purchase cost, Cdw is the fast charging cost, and Caw is the cost of the customer's waiting time.

As a result, the costs constituting the objective function (C) can all be given as a function of the number of chargers (n), and the total cost can be optimized by calculating the social input cost, that is, the number of chargers (n) that minimizes the total cost. there is.

Particle Swarm Optimization (PSO) may be used as a method of calculating the number n of chargers that optimize the objective function C.

In computational science, particle swarm optimization can be a computational method that optimizes a problem by iteratively improving candidate solutions with respect to a given quality measure. Cluster-based optimization is a methodology of mathematical optimization. In cluster-based optimization, several optimizers can exchange information with each other and optimize simultaneously. In addition to the given rules, the optimal solution can be found by combining the information stored by the agents. This swarm-based optimization has the disadvantage of a large amount of computation, but since each agent performs optimization while exchanging information with each other, even if one agent falls into a local solution, it may converge to the global solution as a whole. .

Thus, particle swarm optimization can be a computational method of optimizing a problem by iteratively improving candidate solutions with respect to a given quality measure, taking a population of particles (candidate solutions) and solving simple mathematical formulas for their positions and velocities. The problem can be solved by moving the particles in the search space according to

Each particle's motion is influenced by its local best position, but also driven by its best position in the search space, and can be updated as other particles find a better position.

When the number of chargers that optimize the objective function C is calculated, the controller 320 may compare and present the total cost of the current number of chargers to the calculated total cost of the optimal number of chargers to the operator of the charging device.

In addition, the controller 320 may calculate the arrival rate λ from the arrival interval distribution transmitted from the vehicle entering vehicle measurement unit 100, and provide the calculated arrival rate λ to the operator, and the operator may manage or operate the vehicle. can be utilized for

In addition, the control unit 320 may calculate the charging rate μ or the utilization rate ρ from the charging time distribution transmitted from the charging time measuring unit 200 . The controller 320 may transmit the calculated charging rate μ or utilization rate ρ to an operator or a charging device user.

The charging device user who receives the calculated charging rate (μ) or utilization rate (ρ) can refer to it in selecting where to receive charging from among charging devices in different regions. The operator who receives the calculated charging rate (μ) or utilization rate (ρ) can use it for management or operation, and the operator can use it to determine if the charger is out of order based on the provided charging rate (μ).

The charging device of the present invention may include at least one of the entering vehicle measurement unit 100, the charging time measurement unit 200, the server 300, and the control unit 320.

The entering vehicle measurement unit 100 may measure information about the arrival of the entering vehicle 30 entering the queue 22 waiting for charging. A camera for observation may be provided in the entering vehicle measuring unit 100 . The entering vehicle measurement unit 100 may transmit arrival-related information including an arrival interval to the server 300 through wireless or wired communication.

The charging time measurement unit 200 may transmit information on the charging time of the electric vehicle 10 being charged in each charger or information on the number n of chargers to the server 300 .

A server 300 that determines electricity distribution and billing for each charging device in different regions may be provided. The server 300 may exchange communication with charging devices in different regions. The server 300 may receive data on the charging time or the number of chargers from the charging time measuring unit 200 of each charging device, and transmit data on the arrival interval from the entering vehicle measuring unit 100 of each charging device. can receive

A controller 320 may be provided in the server 300 . The controller 320 may configure the objective function C using data on the arrival interval, charging time, or number of chargers input to the server 300 .

The objective function (C) may be a design cost or management cost input from the manager side of the charging device. Objective function C may be set to maximize the benefit of the operator of the charging device in terms of design, operation or deployment of the charging device.

In addition, the objective function (C) calculates the time required for charging (or charging rate (μ)) available from the user side of the charging device, the length of the queue (Lq), or the waiting time (Wq) for the charging device. can include Therefore, a user who intends to use a charging device in each region can efficiently select a charging device by transmitting/receiving information calculated by the objective function C of the controller 320 from the server 300 .

The charging device of the present invention may include a charger in which the electric vehicle 10 can be charged or a queue 22 in which the electric vehicle 20 waits to enter the charger. The electric vehicles 10 , 20 , and 30 may include an electric vehicle purely powered by a motor and a hybrid vehicle powered by both an engine and a motor.

When the vehicle is being charged in all the chargers provided in the charging device, the vehicle can wait in the queue 22 to wait for the order, and when charging of any one of the plurality of chargers is completed, the vehicle 20 waiting in the queue 22 In the order in which they entered the queue 22, they can be charged from empty chargers.

It can be assumed that the vehicle 30 entering the charging device from the outside first passes through the queue 22 and charges in one of the chargers. An electric vehicle arriving at a charging location out of the queue 22 may be connected to a charger through the charging line 12 to be charged.

Although the case of one queue is shown, the following content may be extended to a plurality of queues 22 . In addition, although the charging device of the present invention will be described below in the case of an electric car, it can be extended to other service types including queues.

Customers' electric vehicles wait in the order they arrive in the queue 22, and when one of the chargers becomes idle, they can receive charging service on a first-in-first-out basis, and if there is no queue, they can receive charging service directly from the charger. It can be assumed that there is

The present invention may obtain or extract the raw data of the charging device to calculate the queue length or waiting time of the queue 22 . It is possible to estimate the average waiting time in queue 22 and the average length of the queue, and this approximation can only give an average value. Raw data may include arrival intervals, charging times, or number of chargers. The number of chargers may be used as raw data as an input parameter constituting the objective function, or the number of chargers may be calculated in an optimization step of the objective function.

Given the distribution of intervals at which customers arrive at a charging station, the distribution of charging times, and the number of EV chargers, the average queue length (number of customers waiting) and average waiting time can be estimated. Each mean or variance may be used for the arrival interval distribution or the charging time distribution. The charging device of the present invention may stochastically derive a waiting time or queue length that optimizes the total cost of the charging device by using the average as the first moment and the variance as the second moment.

An input parameter or an output parameter input to the objective function (C) of the present invention may mean an average value unless otherwise specified.

Therefore, the present invention can maximize the profit of the charging station operator by estimating the number of chargers that minimizes the customer's waiting time and the number of idle chargers.

10... vehicle being charged
12... Charging line
20... waiting vehicle
22... queue
30... queue entry vehicle
100 ... entrance vehicle measuring unit
200 ... charging time measurement unit
300... server
320 ... control part
S10 ... data collection step
S100 ... raw data acquisition step
S110 ... Extract distribution of arrival intervals or filling times
S120... Enter the current number of chargers
S130 ... calculation of arrival rate or filling rate
S200 ... objective function setting step
S210 ... Calculation of utilization rate or idle rate
S220 ... Queue length calculation
S230 ... Waiting time calculation
S240... Including Incidental Expenses
S250... count customer churn rate
S300 ... objective function optimization step
S310... Optimum number of chargers
C... social input costs
C _fc ... charging station investment cost
C _ps ... labor cost
C _dm ... the cost of maintaining a charging station
C _bm ... cost of charger failure
C _be ... customer's purchase cost
C _dw ... fast charge cost
C _aw ... Waiting time of customers
F... Customer patience time or customer churn rate

Claims (8)

A queue is formed in which the customer's electric vehicle waits before charging,
a plurality of chargers capable of charging the electric vehicle;
a server that receives data on a distribution of arrival intervals of electric vehicles entering the queue and data on a distribution of charging times in which the electric vehicles are charged at a charger;
a control unit that sets an objective function, which is the total cost invested in design or operation, using data received from the server; including,
The arrival interval distribution follows an arbitrary distribution that is independently and identically, and the charging time distribution follows an independent phase-type distribution,
The control unit considers customer patience time following an arbitrary distribution that is independently and identically equal when setting the objective function,
The customer patience time refers to a queue departure rate of customers who give up charging without charging the customer's electric vehicle when the waiting time of the customer's electric vehicle waiting in the queue exceeds the customer patience time,
Minimization of the objective function is equal to minimization of the total cost of operating a charger, minimization of the objective function is equal to minimization of waiting time in a queue,
The control unit calculates an arrival rate and a charging rate using the arrival interval distribution, charging time distribution, and the number of chargers transmitted to the server,
The waiting time of the queue is obtained from the arrival rate, filling rate, and queue length;
The waiting time of the customer is a function of the length of the queue, the length of the queue reflecting the customer patience time;
The customer's waiting time and the length of the queue are determined by the number of chargers,
The waiting time of the customer is obtained from the number of chargers and the customer patience time,
The controller calculates the number of electric vehicle chargers that minimize the objective function using particle swarm optimization (PSO).
According to claim 1,
The waiting time of the queue is obtained from the queue length (Lq), arrival rate (λ), filling rate (μ), utilization rate (ρ) and idle rate (1-ρ),
The utilization rate (ρ) is the utilization rate for the charger, the rate at which the charger is in use,
The idle rate (1-ρ) is the rate at which the charger is not used.
delete According to claim 1,
An entry vehicle measurement unit that observes an entry vehicle entering the queue and transmits an arrival interval distribution of the entry vehicle to the server through wired or wireless communication, and a charging time for the electric vehicle to be charged in each of the chargers and measuring the charging time. A charging device including a charging time measurement unit for transmitting time distribution data and the number of chargers to a server.
According to claim 1,
The objective function (C) is given by the following equation as a social input cost,

The objective function includes charging station investment cost (Cfc), labor cost (Cps), charging station maintenance cost (Cdc), charger failure cost (Cbm), customer purchase cost (Cbe), rapid charging cost (Cdw), or customer waiting time. At least one of (Caw) is included,
The charging station investment cost (Cfc), labor cost (Cps), charging station maintenance cost (Cdc), charger failure cost (Cbm), customer purchase cost (Cbe), fast charging cost (Cdw), or customer waiting time (Caw) are all charging devices calculated by the number of chargers (n).
delete delete According to claim 1,
The control unit calculates the utilization rate, which is the rate at which the charger is in use, from the arrival interval distribution transmitted from the entering vehicle measuring unit or the charging time distribution transmitted from the charging time measuring unit,
The arrival rate is provided to the operator terminal by the server and used for operation or management, and the charging rate or utilization rate is transmitted by the server to the electric vehicle user terminal and used to select one of charging devices in different regions,
A charging device in which the total cost for the current number of chargers and the total cost for the calculated optimal number of chargers are compared and presented, and the charging rate or utilization rate is transmitted by the server to an electric vehicle operator terminal and used to determine a faulty charger.

Images