CN114285033A - Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle - Google Patents

Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle Download PDF

Info

Publication number
CN114285033A
CN114285033A CN202111624560.0A CN202111624560A CN114285033A CN 114285033 A CN114285033 A CN 114285033A CN 202111624560 A CN202111624560 A CN 202111624560A CN 114285033 A CN114285033 A CN 114285033A
Authority
CN
China
Prior art keywords
charging
time
load
vehicle
arrival
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111624560.0A
Other languages
Chinese (zh)
Inventor
丁研
朱燕
鄢睿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tianjin University
Original Assignee
Tianjin University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Tianjin University filed Critical Tianjin University
Priority to CN202111624560.0A priority Critical patent/CN114285033A/en
Publication of CN114285033A publication Critical patent/CN114285033A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
    • Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/70—Energy storage systems for electromobility, e.g. batteries
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/60—Other road transportation technologies with climate change mitigation effect
    • Y02T10/7072—Electromobility specific charging systems or methods for batteries, ultracapacitors, supercapacitors or double-layer capacitors
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
    • Y02T90/10—Technologies relating to charging of electric vehicles
    • Y02T90/12—Electric charging stations
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
    • Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
    • Y04S10/00—Systems supporting electrical power generation, transmission or distribution
    • Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications

Landscapes

  • Charge And Discharge Circuits For Batteries Or The Like (AREA)

Abstract

The invention discloses a building energy optimization scheduling method considering uncertainty of charging load of an electric vehicle. The method comprises the following steps: establishing a day-ahead prediction model of the charging load of the electric automobile; quantifying charging load uncertainty on the basis of a charging load prediction model; and establishing a building energy optimization scheduling model considering charging load uncertainty. The strategy considering the uncertainty of the charging load of the electric automobile has higher adaptation degree to various load conditions which may occur in the next day, so that the robustness of the scheduling strategy can be improved; meanwhile, the air conditioning load and the charging load are jointly optimized and scheduled, so that complementarity on a scheduling time period is realized, the load variance is further optimized, and the operation of a building energy system is more stable.

Description

Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle
Technical Field
The invention belongs to the field of building comprehensive energy systems, and particularly relates to a building energy optimization scheduling method considering uncertainty of charging load of an electric vehicle.
Background
With the popularization of charging facilities in buildings in the future, the interaction between the electric automobile and the buildings is gradually close.
The access of the electric automobile has advantages and disadvantages, and the excessive centralized access of the electric automobile can cause the problems of load fluctuation, unbalanced capacity, voltage and frequency of components of a power distribution system, power loss, instability of a power distribution network and the like. However, if the charging scheduling of the electric automobile is combined with the electric power market operation, the charging of the electric automobile is reasonably arranged, so that the fluctuation of the power grid can be restrained through demand response, and the running stability of the power grid can be ensured. The orderly charging of the electric vehicle helps to shift or cut peak loads and reduce corresponding electricity charges. Through the dispatching of the building energy management system, peak clipping and valley filling can be carried out on building loads, and the purpose of improving the on-site consumption condition of new energy is achieved through coordinating the operation of the electric automobile and the distributed power supply. Therefore, the optimal scheduling strategy is formulated by comprehensively utilizing flexible loads such as charging loads and the like, and considerable improvement is brought to the building running state.
However, at present, the strategy established based on the deterministic model often ignores the influence caused by uncertain factors in reality, and the strategy is probably not suitable for practical situations. Therefore, under the condition of rapid development of the electric automobile, a building comprehensive scheduling strategy optimization method under the uncertain condition of the charging load of the electric automobile is necessary to be researched, has theoretical and practical double values, and provides theoretical basis for robust operation management of a building energy system, so that the method conforms to the large popularization trend of the electric automobile, responds to the double-carbon target and promotes the solution of energy and environmental problems.
Disclosure of Invention
In view of the above, the invention provides a building energy optimization scheduling method considering uncertainty of a charging load of an electric vehicle, so as to further improve robustness of a scheduling strategy, and simultaneously optimize load variance by utilizing complementarity of air conditioning load and charging load joint optimization scheduling on a scheduling time period, so that operation of a building energy system is more stable.
In order to achieve the purpose, the invention provides a building energy optimization scheduling method considering uncertainty of charging load of an electric vehicle, which comprises the following steps:
step 1): establishing a day-ahead prediction model of charging load of electric automobile
And (4) performing questionnaire investigation on the local area in China to obtain the actual load data of the electric automobile in China. And analyzing the user commuting behavior characteristics and the physical attribute distribution of the electric automobile according to the recovered data, and establishing an electric automobile charging load prediction model based on a Monte Carlo simulation and a statistical method.
Step 2): quantifying charging load uncertainty
On the basis of a charging load prediction model, errors are introduced through a Monte Carlo method, correction is carried out on the basis of expected values of all parameters, charging load samples of the next day obtained after Monte Carlo simulation is carried out for enough times are combined into a set, and a set of all charging situations which can possibly occur in the next day is formed. The building air conditioning load under the precooling working condition and basic electrical loads such as illumination in a building, office equipment, an elevator and the like are obtained through measurement, and a set of all electrical load conditions which may appear the next day is obtained by combining with the charging load.
Step 3): building energy optimization scheduling model considering charging load uncertainty
And (3) adopting a single-target genetic algorithm, optimizing the most unfavorable condition of the load variance in all possible situations as a target, taking the charging starting time of each electric vehicle as a variable, and proposing the arrival rate and the charging rate as constraint conditions to establish an electric vehicle charging facility optimization scheduling model.
The method comprises the following steps of 1) establishing a day-ahead prediction model of the charging load of the electric automobile, wherein the day-ahead prediction model specifically comprises the following steps:
(1) by investigating the questionnaire, the distribution of the relevant parameters is obtained
And (4) performing questionnaire investigation on the local area in China to obtain the actual load data of the electric automobile in China. By analyzing the questionnaire survey results, a probability density distribution function including indices such as arrival time at the work site, departure time from the work site, travel distance from home to the work site, SOC at the departure time, battery capacity, and power consumption per kilometer is obtained.
(2) Calculation of time-by-time charging load of single trolley
To simplify the calculation of the charging load, the slow charging power is set to 7kW according to the national standard, assuming that the electric vehicle does not leave during operation and that the charging process is considered as constant power. The invention only establishes a full slow charging scene for the configuration of the charging pile type.
After samples of random variables are extracted according to Monte Carlo, the SOC of each electric vehicle when arriving at the station can be calculated:
Figure BSA0000262133290000021
in the formula, SOCarr，nThe charge state of the nth vehicle when the nth vehicle arrives at the working site; SOCde，nThe state of charge of the nth vehicle when the nth vehicle leaves the house; knThe power consumption of the nth vehicle per kilometer is kWh/km; dnThe driving distance, km, from the home to the working place of the nth vehicle; capnBattery capacity of nth vehicle, kWh.
The charging time period of each vehicle can thus be calculated:
Figure BSA0000262133290000022
in the formula, LnThe charging time of the nth vehicle is hour; pc，nThe charging power of the nth vehicle, kW. Under the full slow charging scene, the charging power of all cars is 7 kW.
Let the charging state variable epsilon of the nth electric vehicle at a certain time t in 1 dayn(t)：
Figure BSA0000262133290000031
In the formula, ts，nTime of charging for nth vehicle, charging time of vehicle epsilonn(t) 1, and ε when not chargedn(t)＝0。
The time-by-time charging load of the single trolley and the multiple trolleys is as follows:
PEV，n(t)＝Pc，n×εn(t) (4)
Figure BSA0000262133290000032
in the formula, PEV，n(t) is the charging load, kW, of the nth vehicle at time t; pEV(t) is the charging load, kW, of all vehicles at time t; n is the number of all vehicles. The charging load of the electric automobile with a certain scale by time within one day can be obtained.
(3) Monte Carlo simulation Process
For the full slow charging scenario, starting from the 1 st vehicle, the charging starting time, the battery capacity of the electric vehicle, the power consumption per kilometer, the distance from home to the working place and the SOC at the departure time are randomly extracted according to the probability density distribution function of each random variable, and the time-by-time charging load of the vehicle is calculated according to the formulas (1) to (5). The remaining vehicles are extracted and calculated in the same way, and the hourly charging loads are accumulated to complete the 1-time Monte Carlo simulation process. The above extraction process was repeated to obtain a sufficient number of samples for 10000 extractions. The monte carlo extraction process in the full slow fill scenario is shown in fig. 2. After extraction is completed according to fig. 2, 10000 groups of hourly electric vehicle charging load sample data can be obtained, and then the maximum value, the minimum value, the median and the like of the electric vehicle charging load at each moment are counted and calculated, so that the distribution characteristics of the electric vehicle load can be obtained.
Wherein, the step 2) quantifies the uncertainty of the charging load, and specifically comprises the following steps:
(1) extraction of expected corrections and errors
For each vehicle, in order to simulate uncertain charging load prediction, errors which may occur due to uncertain factors are extracted, expected values of the factors are corrected, and a set of conditions needing to be considered is formed.
The arrival time considering uncertainty deviation is expressed as:
Figure BSA0000262133290000033
in the formula, ts，n(exp) is an expected value of the arrival time (charging start time) of the nth vehicle;
Figure BSA0000262133290000034
is the predicted error value at the time of arrival. The present invention uses the monte carlo method to generate a finite but large enough set of error vectors:
Figure BSA0000262133290000035
in the formula (I), the compound is shown in the specification,
Figure BSA0000262133290000036
the error vector of the predicted value at the arrival time contains all possible prediction errors;
Figure BSA0000262133290000037
error observation vectors extracted by Monte Carlo simulation;
Figure BSA0000262133290000038
predicting a value of the arrival time of the ith extraction; ne is the number of monte carlo draws. When Ne is sufficiently large, it is considered that
Figure BSA0000262133290000039
Is that
Figure BSA00002621332900000310
A reasonably good approximation is to make Ne 10000. And the prediction error follows a normal distribution:
Figure BSA0000262133290000041
in the formula (I), the compound is shown in the specification,
Figure BSA0000262133290000042
the standard deviation of the prediction error for the arrival time;
Figure BSA0000262133290000043
the expectation of the prediction error for the time of arrival.
Similarly, the departure time considering uncertainty deviation is expressed as:
Figure BSA0000262133290000044
Figure BSA0000262133290000045
Figure BSA0000262133290000046
in the formula, tl，n(exp) is an expected value of the nth vehicle departure time (latest time of ending charging);
Figure BSA0000262133290000047
is the predicted error value at the departure time;
Figure BSA0000262133290000048
the error vector of the predicted value at the departure time contains all possible prediction errors;
Figure BSA0000262133290000049
error observation vectors extracted by Monte Carlo simulation;
Figure BSA00002621332900000410
predicting a departure time value extracted for the ith time;
Figure BSA00002621332900000411
a probability density distribution function for the prediction error at the departure time;
Figure BSA00002621332900000412
the standard deviation of the prediction error for the departure time;
Figure BSA00002621332900000413
the expectation of the prediction error for the departure time.
Similarly, the expression for the time-of-arrival SOC considering uncertainty is:
Figure BSA00002621332900000414
Figure BSA00002621332900000415
Figure BSA00002621332900000416
in the formula, SOCarr，n(exp) is the expected value of SOC at the arrival of the nth vehicle;
Figure BSA00002621332900000417
is the predicted error value of the SOC at the time of arrival;
Figure BSA00002621332900000418
the error vector of the SOC predicted value at the time of arrival comprises all possible prediction errors;
Figure BSA00002621332900000419
error observation vectors extracted by Monte Carlo simulation;
Figure BSA00002621332900000420
the predicted value of the time-of-arrival SOC extracted for the ith time is obtained;
Figure BSA00002621332900000421
a probability density distribution function for the SOC prediction error at arrival;
Figure BSA00002621332900000422
the standard deviation of the SOC prediction error at arrival;
Figure BSA00002621332900000423
the expectation of the SOC prediction error at arrival time.
(2) Forming a set of all possible charging loads
After the calculation expression considering the uncertainty of each variable is obtained, according to the Monte Carlo extraction method, the parameters of all vehicles are extracted each time, the result obtained by each extraction is used as a group of possible conditions on the next day, and the extraction is repeated 10000 times. Through the calculation of the charging load, 10000 time-by-time charging load prediction results can be obtained, and the set is used as all charging load curves which may appear in the next day, as shown in the following formula:
Figure BSA0000262133290000051
Figure BSA0000262133290000052
in the formula, omegaEVThe charging load vector comprises all possible situations in the next day;
Figure BSA0000262133290000053
the charge load observation vector extracted by Monte Carlo simulation is obtained;
Figure BSA0000262133290000054
charging loads for the ith group of electric vehicles in one day;
Figure BSA0000262133290000055
and charging the electric automobile at the ith group t.
(3) Forming a set of all possible total electrical loads
Figure BSA0000262133290000056
Ps(t)＝A×I(t)×ηd (18)
Figure BSA0000262133290000057
Figure BSA0000262133290000058
In the formula (I), the compound is shown in the specification,
Figure BSA0000262133290000059
the ith group of net power load at the moment t, kW; pAC(t) is the air conditioning load at time t, kW; pother(t) is the basic power load of other buildings at the time t, including lighting, equipment, elevators and the like, kW;
Figure BSA00002621332900000510
is the average of the i group total power load, kW; ps(t) is the photovoltaic output at time t, and the value is 0 kW in the absence of photovoltaic; a is the area of the photovoltaic panel, m2(ii) a I (t) is the total radiation intensity on the photovoltaic panel at time t, W/m2；ηdThe photovoltaic conversion efficiency.
The building energy optimization scheduling model considering the charging load uncertainty is established in the step 3), and specifically comprises the following steps:
(1) determining optimization variables
In the operation stage, the uncertainty sources of the charging load of the electric automobile are the arrival time and the arrival time SOC. In order to ensure the stable operation of a building energy system and introduce uncertainty into the charging load of the electric automobile, the charging starting time of each electric automobile on the next day is used as an optimization variable, and the charging starting time of each electric automobile on the next day is determined according to the prediction result of the charging load on the next day.
(2) Determining an objective function
Since excessive load fluctuation will seriously affect the operation stability and safety of the power grid, the invention takes the load variance of the total power load as an optimization target. In the case of uncertaintyUnder the circumstances, various charging load curves may appear the next day. Therefore, the final strategy obtained by optimization should be as obvious as possible to optimize all possible situations on the second day, namely, omegaEVIn any case, the load variance should be improved after using the optimization strategy. The objective function is determined as the minimum of the maximum load variance, and the calculation formula is as follows:
Figure BSA0000262133290000061
in the formula, Object is an objective function, meaning that the maximum value (worst case) among the load variances in all scenes is minimized;
Figure BSA0000262133290000062
the ith group of net power load at the moment t, kW;
Figure BSA0000262133290000063
is the average of the i-th group total power load, kW.
(3) Determining constraints
Constraint of initial charging time of electric automobile
The earliest charging starting time of a single electric vehicle is the time when the single electric vehicle arrives, and the SOC of the single electric vehicle is more than or equal to 0.8 when the vehicle leaves, so that the charging requirement of the vehicle owner is met, and the time required by adding the SOC of the single electric vehicle to 0.8 at the latest charging starting time cannot be later than the leaving time of the vehicle.
Figure BSA0000262133290000064
In the formula (I), the compound is shown in the specification,
Figure BSA0000262133290000065
is the minimum value in 10000 groups of n vehicles in the arrival time; t is tres，nThe charging starting time after the nth vehicle is optimized;
Figure BSA0000262133290000066
the time length required when the nth vehicle in the ith group is charged to the SOC of 0.8;
Figure BSA0000262133290000067
is the maximum value of the nth latest charging start time.
② destination uncertainty constraint
Since the arrival time of each vehicle on the next day is various, if the charging start time of a certain vehicle calculated on the next day is earlier than the arrival time of the vehicle, the obtained strategy is considered to be invalid for the vehicle, and the situation that the charging start time is appointed for the vehicle on site after the vehicle arrives needs to be repeated, so that the effectiveness of the day-ahead optimal scheduling is reduced. In order to ensure the robustness of the obtained strategy and adapt to various situations, the concept of the station non-guaranteed rate is proposed:
Figure BSA0000262133290000068
in the formula (I), the compound is shown in the specification,
Figure BSA0000262133290000069
the rate is not guaranteed when the station arrives; n is a radical ofnotarriveFor all scenarios, the arrival time is later than the total number of vehicles for which the charging start time is newly established.
It is specified that the arrival uncertainty rate must not exceed 10%, i.e. the strategy formulated is valid for 90% of vehicles in all scenarios. Since this unwarranted arrival is very likely to occur, if the direct use of the deterministic range hard constraints may make the optimization calculations impossible, soft constraints are set, penalizing the objective function when the conditions are not met:
Figure BSA0000262133290000071
wherein is e1The relaxation variable is greater than or equal to 0.
Third, charge non-guaranteed rate constraint
If a certain vehicle is charged according to the charging start time reset by the strategy and the SOC is not up to 0.8 when the vehicle leaves, the strategy is considered not to meet the charging requirement of the vehicle owner, and the situation is avoided as much as possible. In order to ensure the robustness of the obtained strategy and ensure that the strategy can meet the charging requirement of a user as far as possible under various scenes, the invention provides a concept of a charging non-guarantee rate:
Figure BSA0000262133290000072
in the formula (I), the compound is shown in the specification,
Figure BSA0000262133290000073
the rate is not guaranteed for charging; n is a radical ofnotfullFor all scenarios, the SOC still does not reach a total number of vehicles of 0.8 at departure.
As with the arrival not guaranteed, the objective function is penalized when the condition is not met:
Figure BSA0000262133290000074
wherein is e2The relaxation variable is greater than or equal to 0.
Advantageous effects
(1) Compared with the traditional deterministic strategy, the strategy considering the uncertainty of the charging load of the electric vehicle has higher adaptation degree to various load conditions which may occur the next day, so that the robustness of the scheduling strategy is improved.
(2) The air conditioner load and the charging load are jointly optimized and scheduled, complementarity on a scheduling time period is realized, energy interaction between a building and an electric vehicle charging facility in a parking lot is optimized, load variance is further optimized, and the operation of a building energy system is more stable.
Drawings
FIG. 1 is a technical flow chart of the present invention;
FIG. 2 is a Monte Carlo extraction flow under a full slow charging scenario;
FIG. 3 is a charging load distribution diagram of an electric vehicle under a full slow charging scenario;
FIG. 4 is a graph of the optimization effect of a strategy to consider uncertainty on load variance for all possible scenarios;
FIG. 5 is the total power load of the building under different strategies;
FIG. 6 shows the variance of the total electrical load of a building under different strategies;
FIG. 7 is a Case a & Case-Ref load comparison.
Detailed Description
The invention is further illustrated by the following specific examples and the accompanying drawings. The examples are for the purpose of better understanding the present invention by those skilled in the art and are not intended to limit the present invention in any way.
As shown in fig. 1, the embodiment provides a building energy optimal scheduling method considering uncertainty of charging load of an electric vehicle, including the following steps:
step 1: establishing a day-ahead prediction model of charging load of electric automobile
Selecting a parking lot of a certain small and medium-sized scientific research office building in Tianjin city as a typical case. The questionnaire survey issues questionnaires by visiting, and is divided into two channels of relatives and friends forwarding filling and forum paid filling on line. The questionnaires are distributed by 500 in total in different channels, 427 effective questionnaires are recovered, and 6 indexes including the time of arriving at the working site, the time of leaving the working site, the travel distance from home to the working site, the SOC at the time of departure, the battery capacity and the probability density distribution function of electricity consumption per kilometer 6 are obtained. The parking lot charging piles are all slow charging piles, 10000 Monte Carlo simulations are carried out according to the charging load calculation method introduced in the foregoing, as shown in FIG. 2, and the charging load distribution situation of the electric vehicle time by time in one day under the situation is obtained and is shown in FIG. 3.
Step 2: quantifying charging load uncertainty
On the basis of a charging load prediction model, errors are introduced through a Monte Carlo method, correction is carried out on the basis of expected values of all parameters, charging load samples of the next day obtained after Monte Carlo simulation is carried out for enough times are combined into a set, and a set of all charging situations which can possibly occur in the next day is formed. The real-time load data of the building are mainly acquired through an energy consumption information monitoring platform of the Internet of things, real-time hourly cooling loads of the building and the itemized power consumption of each device (building air conditioning loads, basic power loads of lighting, office equipment, elevators and the like in the building) are monitored, data in 2 months in summer are acquired, and a set of all power load conditions which may appear the next day is acquired by combining with charging loads.
And step 3: building energy optimization scheduling model considering charging load uncertainty
And performing optimization calculation on the operation of the building energy system by using various different scheduling strategies, and setting non-optimization, charging load-only optimization scheduling and air conditioning load and charging load combined optimization scheduling for the case building, wherein the conditions are shown in table 1.
TABLE 1 Integrated optimal scheduling policy Table
Figure BSA0000262133290000081
(1) Arrival and fullness uncertainty
Scheduling is performed by using Case a firstly, so as to compare the optimization effects of the deterministic optimal scheduling strategy and the optimal scheduling strategy considering the uncertainty of the charging load. In order to test the adaptation degree of the strategy obtained by the deterministic method to all possible situations in the next day, the strategy is substituted into 10000 groups of situations, the load variance under the worst situation, the arrival rate and the full rate are calculated, and the result is compared with the result of the optimal scheduling strategy considering the charging load uncertainty. The specific calculation results are shown in table 2.
TABLE 2 deterministic to uncertainty optimized scheduling policy result comparison
Figure BSA0000262133290000091
As can be seen from table 2, the arrival and charge non-guaranteed rates of the deterministic policy are both greater than the prescribed upper limit of 10%. When using a deterministic strategy, the scheduled start-of-charge time will fail 12.23% of the vehicles and 24.03% of the vehicles will not meet the charge demand upon departure in the face of the various conditions that may occur the next day. The arrival uncertainty and the charging uncertainty of the strategy considering the uncertainty of the charging load of the electric vehicle are both obviously reduced and can be kept within 10 percent, and the adaptation degree of the strategy considering the uncertainty to various conditions which may appear in the next day is far higher than that of the deterministic strategy.
(2) Load variance optimization
The load variance improvement statistics for various possible occurrences for the strategy that takes into account the charging load uncertainty is shown in fig. 4. The uncertainty-considering strategy may optimize all possible scenarios, where the load variance for the vast majority of scenarios may be reduced by 5% -10%. The uncertainty-considering strategy provided by the invention can effectively optimize the load variance of any possible situation on the premise of ensuring that the strategy can be effectively implemented on more than 90% of vehicles and ensuring that more than 90% of charging requirements of the vehicles can be met. Compared with a deterministic strategy, the strategy provided by the invention considering the uncertainty of the charging load has more excellent robustness.
Ordered charging of EV
The total power load curve of the building for each strategy is shown in fig. 5, and the load variance for each strategy is shown in fig. 6. First, based on Case a and Case-Ref in fig. 5 and 6, it can be seen that orderly charging of electric vehicles can reduce the load variance by 2.83% and significantly reduce the peak time 9: 00 and 10: 00, the negative effect of 'peak-to-peak' after the electric automobile is connected is reduced. The main reason for this is that the charging load is shifted by newly determining the charging start time of the electric vehicle, and the specific charging load is as shown in fig. 7. When the orderly charging is not performed, the electric vehicle load is concentrated on 8: 00-14: 00, especially in the morning 9: 00-11: 00, coinciding with the peak period. After the orderly charging, the charging load of the electric automobile is uniformly distributed in the following steps of 9: 00-19: 00, the charging load in the morning peak period is shifted to 14: 00-18: a flat period of 00. And when the peak load transfer is realized, the load is smoother, and the running stability of the system is improved.
As can also be seen from fig. 7, the period of time that the electric vehicle can be adjusted is limited to 8 due to the influence of the commute behavior of the vehicle owner: 00-19: between 00, although the peak load can be shifted to the flat section, the valley section is still in an unutilized state, and the optimization of the system load variance still has great potential.
Second, EV orderly charging and pre-cooling in advance
Further comparing Case a and Case b in fig. 5 and 6, it can be seen that air-conditioning precooling will be used during peak hours 8: the load of 00 is transferred to the valley section at night, the air conditioner load which is suddenly increased when the electric vehicle is started in the morning is obviously reduced, the load variance is reduced by 3.06 percent to 17.81 percent compared with Case-Ref, and the optimization effect is better than that of Case a which only carries out ordered charging of the electric vehicle. The air conditioner load and the electric vehicle charging load have good complementarity in scheduling time, and are matched to carry out demand side load management, so that the load variance can be further optimized, the operation of a building energy system is more stable, more loads can be transferred to valley sections, and peak clipping and valley filling to a greater degree are realized.
In addition, comparing Case b-2h to Case b-6h in fig. 6, it can be found that the earlier the pre-boot pre-cooling time is, the smaller the load variance of the strategy is, and the better the stability of the system operation is. The reason is that the earlier the air conditioner is started, the longer the running time of the air conditioner at night is, the larger the load of the air conditioner at night is, the difference between the load of the air conditioner at day and night is reduced, and the load variance is reduced.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are only exemplary embodiments of the present invention and are not intended to limit the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (4)

1. A building energy optimization scheduling method considering uncertainty of charging load of an electric automobile is characterized by comprising the following steps:
step 1): establishing a day-ahead prediction model of charging load of electric automobile
And (4) performing questionnaire investigation on the local area in China to obtain the actual load data of the electric automobile in China. And analyzing the user commuting behavior characteristics and the physical attribute distribution of the electric automobile according to the recovered data, and establishing an electric automobile charging load prediction model based on a Monte Carlo simulation and a statistical method.
Step 2): quantifying charging load uncertainty
On the basis of a charging load prediction model, errors are introduced through a Monte Carlo method, correction is carried out on the basis of expected values of all parameters, charging load samples of the next day obtained after Monte Carlo simulation is carried out for enough times are combined into a set, and a set of all charging situations which can possibly occur in the next day is formed. The building air conditioning load under the precooling working condition and basic electrical loads such as illumination in a building, office equipment, an elevator and the like are obtained through measurement, and a set of all electrical load conditions which may appear the next day is obtained by combining with the charging load.
Step 3): building energy optimization scheduling model considering charging load uncertainty
And (3) adopting a single-target genetic algorithm, optimizing the most unfavorable condition of the load variance in all possible situations as a target, taking the charging starting time of each electric vehicle as a variable, and proposing the arrival rate and the charging rate as constraint conditions to establish an electric vehicle charging facility optimization scheduling model.
2. The computing method according to claim 1, characterized in that: the method comprises the following steps of 1) establishing a day-ahead prediction model of the charging load of the electric automobile, wherein the day-ahead prediction model specifically comprises the following steps:
(1) by investigating the questionnaire, the distribution of the relevant parameters is obtained
In order to understand the commuting travel law of the native people in china, the questionnaire content relates to basic information of people (sex, age), travel habits of people (time to arrive at a work site, time to leave the work site, distance from home to the work site, vehicle SOC at the time of departure, whether the people will drive away during work), physical parameters of the electric vehicle (battery capacity) and willingness to charge. Through analysis of the questionnaire survey results, 6 indexes including arrival time at the work site, departure time from the work site, travel distance from home to the work site, SOC at the departure time, battery capacity, and probability density distribution function of power consumption per kilometer were obtained.
(2) Calculation of time-by-time charging load of single trolley
To simplify the calculation of the charging load, the conditions are assumed and defined to some extent, as follows:
the electric automobile cannot leave during work and is parked until work leaving time;
charging the electric automobile when the electric automobile arrives at the station, finishing charging until the battery is fully charged, and charging until the electric automobile leaves if the electric automobile cannot be fully charged until the electric automobile leaves;
and thirdly, since the duration of the charging starting process and the charging ending process is short, the charging process is regarded as constant power, and the slow charging power is regulated to be 7kW according to the national standard.
And fourthly, only establishing a full slow charging scene by the charging pile type configuration.
After samples of random variables are extracted according to Monte Carlo, the SOC of each electric vehicle when arriving at the station can be calculated:
Figure FSA0000262133280000021
in the formula, SOCarr，nThe charge state of the nth vehicle when the nth vehicle arrives at the working site; SOCde，nThe state of charge of the nth vehicle when the nth vehicle leaves the house; knThe power consumption of the nth vehicle per kilometer is kWh/km; dnThe driving distance, km, from the home to the working place of the nth vehicle; capnBattery capacity of nth vehicle, kWh.
The charging time period of each vehicle can thus be calculated:
Figure FSA0000262133280000022
in the formula, LnThe charging time of the nth vehicle is hour; pc，nThe charging power of the nth vehicle, kW. Under the full slow charging scene, the charging power of all cars is 7 kW.
Let the charging state variable epsilon of the nth electric vehicle at a certain time t in 1 dayn(t)：
Figure FSA0000262133280000023
In the formula, ts，nTime of charging for nth vehicle, charging time of vehicle epsilonn(t) 1, and ε when not chargedn(t)＝0。
The time-by-time charging load of the single trolley and the multiple trolleys is as follows:
PEV，n(t)＝Pc，n×εn(t) (4)
Figure FSA0000262133280000024
in the formula, PEV，n(t) is the charging load, kW, of the nth vehicle at time t; pEV(t) is the charging load, kW, of all vehicles at time t; n is the number of all vehicles. The charging load of the electric automobile with a certain scale by time within one day can be obtained.
(3) Monte Carlo simulation Process
For the full slow charging scenario, starting from the 1 st vehicle, the charging starting time, the battery capacity of the electric vehicle, the power consumption per kilometer, the distance from home to the working place and the SOC at the departure time are randomly extracted according to the probability density distribution function of each random variable, and the time-by-time charging load of the vehicle is calculated according to the formulas (1) to (5). The remaining vehicles are extracted and calculated in the same way, and the hourly charging loads are accumulated to complete the 1-time Monte Carlo simulation process. The above extraction procedure was repeated to obtain a sufficient number of samples, which were extracted a total of 10000 times. 10000 groups of hourly electric vehicle charging load sample data can be obtained after extraction is finished, and then the maximum value, the minimum value, the median and the like of the electric vehicle charging load at each moment are counted and calculated, so that the distribution characteristics of the electric vehicle load can be obtained.
3. The computing method according to claim 1, characterized in that: the step 2) quantifies the uncertainty of the charging load, and specifically comprises the following steps:
(1) extraction of expected corrections and errors
For each vehicle, in order to simulate uncertain charging load prediction, errors which may occur due to uncertain factors are extracted, expected values of the factors are corrected, and a set of conditions needing to be considered is formed.
The arrival time considering uncertainty deviation is expressed as:
Figure FSA0000262133280000031
in the formula, ts，n(exp) is an expected value of the arrival time (charging start time) of the nth vehicle;
Figure FSA0000262133280000032
is the predicted error value at the time of arrival. The present invention uses the monte carlo method to generate a finite but large enough set of error vectors:
Figure FSA0000262133280000033
in the formula (I), the compound is shown in the specification,
Figure FSA0000262133280000034
the error vector of the predicted value at the arrival time contains all possible prediction errors;
Figure FSA0000262133280000035
error observation vectors extracted by Monte Carlo simulation;
Figure FSA0000262133280000036
predicting a value of the arrival time of the ith extraction; ne is the number of monte carlo draws. When Ne is sufficiently large, it is considered that
Figure FSA0000262133280000037
Is that
Figure FSA0000262133280000038
A reasonably good approximation is to make Ne 10000. And the prediction error follows a normal distribution:
Figure FSA0000262133280000039
in the formula (I), the compound is shown in the specification,
Figure FSA00002621332800000310
the standard deviation of the prediction error for the arrival time;
Figure FSA00002621332800000311
the expectation of the prediction error for the time of arrival.
Similarly, the departure time considering uncertainty deviation is expressed as:
Figure FSA00002621332800000312
Figure FSA00002621332800000313
Figure FSA00002621332800000314
in the formula, tl，n(exp) is an expected value of the nth vehicle departure time (latest time of ending charging);
Figure FSA00002621332800000315
is the predicted error value at the departure time;
Figure FSA00002621332800000316
the error vector of the predicted value at the departure time contains all possible prediction errors;
Figure FSA00002621332800000317
error observation vectors extracted by Monte Carlo simulation;
Figure FSA00002621332800000318
predicting a departure time value extracted for the ith time;
Figure FSA00002621332800000319
a probability density distribution function for the prediction error at the departure time;
Figure FSA00002621332800000320
the standard deviation of the prediction error for the departure time;
Figure FSA00002621332800000321
the expectation of the prediction error for the departure time.
Similarly, the expression for the time-of-arrival SOC considering uncertainty is:
Figure FSA0000262133280000041
Figure FSA0000262133280000042
Figure FSA0000262133280000043
in the formula, SOCarr，n(exp) is the expected value of SOC at the arrival of the nth vehicle;
Figure FSA0000262133280000044
is the predicted error value of the SOC at the time of arrival;
Figure FSA0000262133280000045
the error vector of the SOC predicted value at the time of arrival comprises all possible prediction errors;
Figure FSA0000262133280000046
error observation vectors extracted by Monte Carlo simulation;
Figure FSA0000262133280000047
the predicted value of the time-of-arrival SOC extracted for the ith time is obtained;
Figure FSA0000262133280000048
a probability density distribution function for the SOC prediction error at arrival;
Figure FSA0000262133280000049
the standard deviation of the SOC prediction error at arrival;
Figure FSA00002621332800000410
the expectation of the SOC prediction error at arrival time.
(2) Forming a set of all possible charging loads
After the calculation expression considering the uncertainty of each variable is obtained, according to the Monte Carlo extraction method, the parameters of all vehicles are extracted each time, the result obtained by each extraction is used as a group of possible conditions on the next day, and the extraction is repeated 10000 times. Through the calculation of the charging load, 10000 time-by-time charging load prediction results can be obtained, and the set is used as all charging load curves which may appear in the next day, as shown in the following formula:
Figure FSA00002621332800000411
Figure FSA00002621332800000412
in the formula, omegaEVThe charging load vector comprises all possible situations in the next day;
Figure FSA00002621332800000413
the charge load observation vector extracted by Monte Carlo simulation is obtained;
Figure FSA00002621332800000414
charging loads for the ith group of electric vehicles in one day;
Figure FSA00002621332800000415
and charging the electric automobile at the ith group t.
(3) Forming a set of all possible total electrical loads
Figure FSA00002621332800000416
Ps(t)＝A×I(t)×ηd (18)
Figure FSA00002621332800000417
Figure FSA0000262133280000051
In the formula (I), the compound is shown in the specification,
Figure FSA0000262133280000052
the ith group of net power load at the moment t, kW; pAC(t) is the air conditioning load at time t, kW; pother(t) is the basic power load of other buildings at the time t, including lighting, equipment, elevators and the like, kW;
Figure FSA0000262133280000053
is the average of the i group total power load, kW; ps(t) is the photovoltaic output at time t, and the value is 0 kW in the absence of photovoltaic; a is the area of the photovoltaic panel, m2(ii) a I (t) is the total radiation intensity on the photovoltaic panel at time t, W/m2；ηdThe photovoltaic conversion efficiency.
4. The computing method according to claim 1, characterized in that: the step 3) of establishing a building energy optimization scheduling model considering charging load uncertainty specifically comprises the following steps:
(1) determining optimization variables
In the operation stage, the uncertainty sources of the charging load of the electric automobile are the arrival time and the arrival time SOC. In order to ensure the stable operation of a building energy system and introduce uncertainty into the charging load of the electric automobile, the charging starting time of each electric automobile on the next day is used as an optimization variable, and the charging starting time of each electric automobile on the next day is determined according to the prediction result of the charging load on the next day.
(2) Determining an objective function
Since excessive load fluctuation will seriously affect the operation stability and safety of the power grid, the invention takes the load variance of the total power load as an optimization target. A wide variety of charge load curves may appear the next day, taking into account uncertainty. Therefore, the final strategy obtained by optimization should be as obvious as possible to optimize all possible situations on the second day, namely, omegaEVWhich one ofIn all cases, the load variance should be improved using the optimization strategy. The objective function is determined as the minimum of the maximum load variance, and the calculation formula is as follows:
Figure FSA0000262133280000054
in the formula, Object is an objective function, meaning that the maximum value (worst case) among the load variances in all scenes is minimized;
Figure FSA0000262133280000055
the ith group of net power load at the moment t, kW;
Figure FSA0000262133280000056
is the average of the i-th group total power load, kW.
(3) Determining constraints
Constraint of initial charging time of electric automobile
The earliest charging starting time of a single electric vehicle is the time when the single electric vehicle arrives, and the SOC of the single electric vehicle is more than or equal to 0.8 when the vehicle leaves, so that the charging requirement of the vehicle owner is met, and the time required by adding the SOC of the single electric vehicle to 0.8 at the latest charging starting time cannot be later than the leaving time of the vehicle.
Figure FSA0000262133280000057
In the formula (I), the compound is shown in the specification,
Figure FSA0000262133280000061
is the minimum value in 10000 groups of n vehicles in the arrival time; t is tres，nThe charging starting time after the nth vehicle is optimized;
Figure FSA0000262133280000062
the time length required when the nth vehicle in the ith group is charged to the SOC of 0.8;
Figure FSA0000262133280000063
is the maximum value of the nth latest charging start time.
② destination uncertainty constraint
Since the arrival time of each vehicle on the next day is various, if the charging start time of a certain vehicle calculated on the next day is earlier than the arrival time of the vehicle, the obtained strategy is considered to be invalid for the vehicle, and the situation that the charging start time is appointed for the vehicle on site after the vehicle arrives needs to be repeated, so that the effectiveness of the day-ahead optimal scheduling is reduced. In order to ensure the robustness of the obtained strategy and adapt to various situations, the concept of the station non-guaranteed rate is proposed:
Figure FSA0000262133280000064
in the formula (I), the compound is shown in the specification,
Figure FSA0000262133280000065
the rate is not guaranteed when the station arrives; n is a radical ofnotarriveFor all scenarios, the arrival time is later than the total number of vehicles for which the charging start time is newly established.
It is specified that the arrival uncertainty rate must not exceed 10%, i.e. the strategy formulated is valid for 90% of vehicles in all scenarios. Since this unwarranted arrival is very likely to occur, if the direct use of the deterministic range hard constraints may make the optimization calculations impossible, soft constraints are set, penalizing the objective function when the conditions are not met:
Figure FSA0000262133280000066
in the formula, epsilon1The relaxation variable is greater than or equal to 0.
Third, charge non-guaranteed rate constraint
If a certain vehicle is charged according to the charging start time reset by the strategy and the SOC is not up to 0.8 when the vehicle leaves, the strategy is considered not to meet the charging requirement of the vehicle owner, and the situation is avoided as much as possible. In order to ensure the robustness of the obtained strategy and ensure that the strategy can meet the charging requirement of a user as far as possible under various scenes, the invention provides a concept of a charging non-guarantee rate:
Figure FSA0000262133280000067
in the formula (I), the compound is shown in the specification,
Figure FSA0000262133280000068
the rate is not guaranteed for charging; n is a radical ofnotfullFor all scenarios, the SOC still does not reach a total number of vehicles of 0.8 at departure.
As with the arrival not guaranteed, the objective function is penalized when the condition is not met:
Figure FSA0000262133280000069
in the formula, epsilon2The relaxation variable is greater than or equal to 0.
CN202111624560.0A 2021-12-29 2021-12-29 Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle Pending CN114285033A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111624560.0A CN114285033A (en) 2021-12-29 2021-12-29 Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111624560.0A CN114285033A (en) 2021-12-29 2021-12-29 Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle

Publications (1)

Publication Number Publication Date
CN114285033A true CN114285033A (en) 2022-04-05

Family

ID=80877467

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111624560.0A Pending CN114285033A (en) 2021-12-29 2021-12-29 Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle

Country Status (1)

Country Link
CN (1) CN114285033A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115840986A (en) * 2023-02-24 2023-03-24 明士新材料有限公司 Energy management method based on stochastic model predictive control

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115840986A (en) * 2023-02-24 2023-03-24 明士新材料有限公司 Energy management method based on stochastic model predictive control
CN115840986B (en) * 2023-02-24 2023-06-27 明士新材料有限公司 Energy management method based on stochastic model predictive control

Similar Documents

Publication Publication Date Title
Wang et al. Demand side management of plug-in electric vehicles and coordinated unit commitment: A novel parallel competitive swarm optimization method
Jian et al. Optimal scheduling for vehicle-to-grid operation with stochastic connection of plug-in electric vehicles to smart grid
CN103241130B (en) Energy management method and system for electric bus charging and swap station
Li et al. Boosting grid efficiency and resiliency by releasing V2G potentiality through a novel rolling prediction-decision framework and deep-LSTM algorithm
CN106532764B (en) A kind of electric car charging load control method of on-site elimination photovoltaic power generation
CN109492791B (en) Inter-city expressway network light storage charging station constant volume planning method based on charging guidance
Yang et al. Optimal dispatching strategy for shared battery station of electric vehicle by divisional battery control
CN109948823B (en) Self-adaptive robust day-ahead optimization scheduling method for light storage charging tower
Yu et al. A real time energy management for EV charging station integrated with local generations and energy storage system
CN116001624A (en) Ordered charging method for one-pile multi-connected electric automobile based on deep reinforcement learning
CN112865190A (en) Optimal scheduling method and system for photovoltaic and charging demand-based optical storage charging station
CN113799640B (en) Energy management method suitable for micro-grid containing electric vehicle charging piles
CN116151486B (en) Multi-time-scale random optimization method and device for photovoltaic charging station with energy storage system
Wang et al. Event-based electric vehicle scheduling considering random user behaviors
Deng et al. Hierarchical operation management of electric vehicles for depots with PV on-site generation
CN114285033A (en) Building energy optimization scheduling method considering uncertainty of charging load of electric vehicle
Ali et al. Optimal operation of static energy storage in fast-charging stations considering the trade-off between resilience and peak shaving
CN110861508B (en) Charging control method and system shared by residential area direct current chargers and storage medium
Ding et al. A comprehensive scheduling model for electric vehicles in office buildings considering the uncertainty of charging load
Wang et al. Optimal planning of renewable generations for electric vehicle charging station
Quan-Do et al. Monitoring of power allocation in centralized electric vehicle charging spot system
Yuan et al. Source: Towards solar-uncertainty-aware e-taxi coordination under dynamic passenger mobility
Ma et al. IMOCS based EV charging station planning optimization considering stakeholders’ interests balance
CN113381406A (en) Electric vehicle charging and discharging control method, device, equipment and storage medium
CN113949082A (en) Day-ahead energy scheduling method and device for industrial park

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination