CN112952847A - Multi-region active power distribution system peak regulation optimization method considering electricity demand elasticity - Google Patents
Multi-region active power distribution system peak regulation optimization method considering electricity demand elasticity Download PDFInfo
- Publication number
- CN112952847A CN112952847A CN202110366783.5A CN202110366783A CN112952847A CN 112952847 A CN112952847 A CN 112952847A CN 202110366783 A CN202110366783 A CN 202110366783A CN 112952847 A CN112952847 A CN 112952847A
- Authority
- CN
- China
- Prior art keywords
- load
- region
- decision time
- power
- area
- 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.)
- Granted
Links
Images
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/12—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load
- H02J3/14—Circuit arrangements for ac mains or ac distribution networks for adjusting voltage in ac networks by changing a characteristic of the network load by switching loads on to, or off from, network, e.g. progressively balanced loading
- H02J3/144—Demand-response operation of the power transmission or distribution network
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/007—Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources
- H02J3/0075—Arrangements for selectively connecting the load or loads to one or several among a plurality of power lines or power sources for providing alternative feeding paths between load and source according to economic or energy efficiency considerations, e.g. economic dispatch
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- 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
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B70/00—Technologies for an efficient end-user side electric power management and consumption
- Y02B70/30—Systems integrating technologies related to power network operation and communication or information technologies for improving the carbon footprint of the management of residential or tertiary loads, i.e. smart grids as climate change mitigation technology in the buildings sector, including also the last stages of power distribution and the control, monitoring or operating management systems at local level
- Y02B70/3225—Demand response systems, e.g. load shedding, peak shaving
-
- 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
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
-
- 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
- Y04S20/00—Management or operation of end-user stationary applications or the last stages of power distribution; Controlling, monitoring or operating thereof
- Y04S20/20—End-user application control systems
- Y04S20/222—Demand response systems, e.g. load shedding, peak shaving
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses a peak regulation optimization method of a multi-region active power distribution system considering power demand elasticity, which comprises the steps of firstly establishing a mathematical model of a photovoltaic and all-vanadium redox flow battery energy storage system and a multi-region flexible load scheduling unit under a power elasticity environment; then, establishing a DTMDP model by considering the peak regulation optimization problem of the multi-region active power distribution system with the elasticity of power demand; and finally, solving the mathematical model by combining reinforcement learning and an intelligent algorithm to obtain a multi-region scheduling optimization control strategy meeting peak regulation requirements. The layered learning mechanism avoids the problem of dimension disaster of reinforcement learning to a certain extent, and promotes the rapid solution of the scheduling strategy; meanwhile, the combination of reinforcement learning and an intelligent algorithm further enhances the exploration capability of the algorithm, and is beneficial to solving the optimal peak regulation strategy; potential scheduling information of the active power distribution system can be further obtained by considering the elasticity of power demand, and smooth and safe operation of the system is promoted.
Description
Technical Field
The invention belongs to the field of dispatching optimization of a multi-region active power distribution system, and particularly relates to a dynamic dispatching optimization method of the multi-region active power distribution system, which takes the peak regulation demand and the power consumption demand elasticity of a power grid into consideration and aims at achieving stable and economic operation of the system.
Background
Currently, the research focus of the active power distribution system includes planning design, hierarchical control, operation management, and the like. The research in the planning aspect mainly develops around the aspects of the optimal configuration of the distributed power supply, the net rack planning and the like; the research of the hierarchical coordination control provides technical support for scheduling and managing various resources, and the overall optimization is achieved by managing the hierarchical distributed energy sources; and the aspects of multi-enclosure reactive power compensation, scheduling optimization and the like are researched in the aspect of operation management.
The traditional power grid peak regulation problem research mainly considers the starting, stopping and output regulation control of a generating side unit, and particularly the combined operation optimization of a multi-energy complementary system containing an energy storage device is proved to be capable of effectively relieving the peak regulation pressure of the system. With the development of an active power distribution system and the application of a flexible load scheduling technology, various types of high-quality peak shaving resources on a demand side are scheduled and optimized, so that economic peak shaving and energy utilization rate improvement are realized, an important trend of current power grid peak shaving research is formed, and the method is also an effective supplement for peak shaving on a power generation side.
The current research is more in the analysis and research of the whole load of the regional power grid, the research on the load characteristics of different industries is less, the influence caused by the difference and the proportional change of the loads of different industries is ignored to a certain extent, and the change rule of the power load is not easy to grasp more accurately. Meanwhile, the electric power of different types of electric users is flexibly depicted, which is not beneficial to guiding electric power users to select reasonable electric power utilization time and adjusting the electric power utilization of the users in the planning process of an electric power system so as to achieve better peak clipping and valley filling effects.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a peak shaving optimization method of a multi-region active power distribution system considering the elasticity of power consumption requirements, so that a power consumer can be guided to select reasonable power consumption time and adjust the power consumption of the power consumer, and the effects of peak shaving and valley filling are achieved, so that the power consumption load rate is improved, and the safety and the stability of power grid operation are improved.
The invention adopts the following technical scheme for solving the technical problems:
the invention relates to a peak regulation optimization method of a multi-region active power distribution system considering power demand elasticity, which is characterized by comprising the following steps of:
the area i comprises: the system comprises an ith PV generating unit, an ith VRB energy storage unit and an ith load scheduling unit; the load type of the ith load scheduling unit comprises: the ith load-reducible load and the ith rigid load; wherein, the load type of the load scheduling unit in the industrial park scheduling area further comprises: the load can be transferred;
Step 3, establishing mathematical models of various load scheduling units, VRB energy storage units and PV power generation units in industrial, commercial, municipal and life parks, a region scheduling time attribute mathematical model and a peak shaving task allocation mechanism considering region elasticity amplitude:
step 3.1, establishing mathematical models of various types of loads of multiple areas under the elastic environment:
obtaining the minimum and maximum load reduction quantity constraints of the region i at the decision time t by using the formula (1):
in the formula (1), the reaction mixture is,
the maximum load amount of the reducible load of the area i at the decision time t is reduced;
actually reducing the load of the area i at the decision time t;
the transferable load force constraint is obtained by using the formula (2) to the formula (4):
in the formulae (2) to (4),
and
respectively determining the maximum allowable load increase and the maximum allowable load decrease of the transferable load in the area i at the decision time t;
the load increment of the transferable load corresponding to the area i at the decision time t;
the load reduction quantity of the area i at the decision time t;
defining the accumulated increase and decrease load quantity after action is taken at each decision time from the initial time to the decision time t as
Suppose a single scheduling day has tK-1At the moment of decision, then
Is the remaining period tK-1-t is a margin of resilience to transferable load;
obtaining margin constraints of the region i at the decision time t by using the equations (5) and (6)
And transfer direction constraints
Thereby obtaining transferable load constraint of the area i at the decision time t
In the formulae (5) and (6), aacctrIs the elastic margin coefficient; e is a natural constant; alpha is alphadirSelecting a direction coefficient for the transfer motion;
the load amount is increased or decreased by the accumulated transfer to the t-1 moment;
when the load is increased and decreased correspondingly after the action is taken for the transferable load at the current decision time t and the load is set to be increased,
the value of (d) is positive, and when the load amount is reduced,
the value of (d) is negative, when not active,
is 0;
3.2, establishing a mathematical model of the output of the VRB energy storage unit:
establishing the constraint condition of the VRB energy storage unit in one scheduling day by using an equation (7) to an equation (10), wherein the constraint condition comprises the following steps: terminal voltage constraint, charge-discharge power constraint, charge state constraint and initial and final charge state consistency constraint:
in the formulae (7) to (10),
the upper limit and the lower limit of the end voltage of the VRB energy storage unit of the area i,
the minimum and maximum charge-discharge power of the VRB energy storage unit in the area i at the decision time t,
the actual charging and discharging power of the VRB energy storage unit at decision time t for region i,
the remaining capacity of the VRB energy storage cells for region i constrains the upper and lower limits,
for the remaining capacity of the VRB energy storage unit at decision time t for region i, ts、teTo schedule the beginning and end of the day, CconSetting the expected value of the state of charge of the VRB energy storage unit;
3.3, establishing a mathematical model of photovoltaic power generation output:
obtaining a predicted value of the photovoltaic output power of the area i at the decision time t by using the formula (11)
In the formula (11), etapvThe photoelectric conversion efficiency; n ispvThe number of the photovoltaic cell panels; spvThe surface area of the photovoltaic cell panel receiving illumination is increased;
the solar radiation intensity of the area i at the decision time t; alpha is alphapvIn order to be the temperature conversion coefficient,
the outdoor temperature of the area i at the decision time t;
step 3.4, establishing a mathematical model of the regional scheduling time attribute:
obtaining the time attribute T of the area i at the decision time T by using the formula (12)i,t：
In the formula (12), the reaction mixture is,
is the time magnitude parameter of the region i; c is a constant; p is a radical ofi,tThe power output at decision time t for region i,
maximum output power of the region i on the dispatching day;
step 3.5, establishing a peak regulation task allocation mechanism considering the elastic amplitude of the region:
setting the elastic amplitude of the area i at the decision time t as Ei,tElastic amplitude Ei,tUpper bound of (2) which can reduce the upper bound of the load
+ transferable increased load upper bound
+ an energy storage discharge margin; elastic amplitude Ei,tLower bound of (2) reducible load lower bound + transferable reduced lower bound
+ an energy storage charging margin; the elastic amplitude of the region i is formed by the span between the upper and lower bounds;
step 4, modeling of continuous variable discretization and uncertain random variable dynamic change processes:
step 4.1, establishing a multi-region power grid peak regulation demand uncertainty model:
at decision time t, issuing the power grid to the maximum range interval of the random uncertain part of the peak regulation demand instruction of the area i in real time
Is dispersed into
In total
A plurality of levels, wherein,
for the maximum value of the upward fluctuation based on the peak shaver demand predicted power for the region i at the decision time t,
the maximum value of downward fluctuation based on the peak load demand prediction power of the area i at the decision time t;
maximum discrete levels of upward and downward fluctuation amounts based on the predicted power of the peak shaving demand of the region i;
and (3) obtaining the actual peak regulation demand of the power grid of the area i at the decision time t by using the formula (13):
in the formula (13), the reaction mixture is,
predicting power for the peak shaver demand of the power grid in the area i at the decision time t,
the power level of the uncertain part of the peak shaving demand of the power grid in the area i at the decision time t,
the minimum unit power of the power grid peak regulation instruction uncertainty part discretization of the area i at the decision time t is obtained;
step 4.2, establishing a photovoltaic output uncertain model;
the maximum range interval of the photovoltaic output uncertain part of the area i at the decision time t
Is dispersed into
In total
A plurality of levels, wherein,
the maximum value of the upward fluctuation based on the photovoltaic output predicted power of the area i at the decision time t,
the maximum value of the downward fluctuation based on the photovoltaic output predicted power of the area i at the decision time t;
is illuminated by an area iMaximum discrete levels of upward and downward fluctuation amounts based on the predicted power of the volt-ampere output;
obtaining the actual photovoltaic output of the region i at the decision time t by using the formula (14)
In the formula (14), the compound represented by the formula (I),
the power is predicted for the photovoltaic output at decision time t for zone i,
the power level of the uncertain photovoltaic output part of the area i at the decision time t is determined;
the minimum unit power of the area i after the photovoltaic output uncertain part is dispersed at decision time t is obtained;
step 4.3, establishing a multi-region uncertain model of each type of load demand:
the maximum range interval of the random uncertain part of the reducible load and the transferable load of the area i at the decision time t
And
respectively dispersed into corresponding state grades
And
in total
And
a plurality of levels, wherein,
and
respectively the maximum value of the upward fluctuation based on the forecasted power of the reducible load and the transferable load demand of the area i at the decision time t,
and
maximum values of downward fluctuation based on demand predicted power of reducible load and transferable load in the area i at decision time t respectively;
and
maximum discrete levels of upward and downward fluctuation amounts based on predicted power required by reducible load and transferable load in the region i respectively;
obtaining the actual power demand of the region i at the decision time t, which can reduce the load, by using the equations (15) to (16)
And the actual power demand of the load can be transferred
In the formulae (15) to (16),
power is predicted for the demand that region i can shed loads and transferable loads at decision time t,
respectively the power levels of the uncertain parts of the reducible load and the transferable load demand power in the area i at the decision time t;
respectively the minimum unit power of the region i after the uncertain part of the reducible load and the transferable load demand power is dispersed under the decision time t;
step 5, establishing a corresponding DTMDP model according to the peak regulation optimization problem of the multi-region active power distribution system considering the elasticity of power consumption requirements:
step 5.1, system state space and action set of the DTMDP model:
dividing a scheduling day into K e {0, 1.,. K-1}, and K decision periods; the time length of each decision period is delta t, and the decision time of the kth decision period is tkThe end time of the scheduling day is tK-1；
The formula (17) -formula (18) is used for obtaining the decision time t of the scheduling centerkState of
In the formulae (17) to (18),
as a decision time tkThe real-time peak regulation demand state grade of the lower power grid,
for region i at decision time tkEnvironmental information of the environment, photovoltaic output state class
VRB energy storage unit charging and discharging state grade
Multi-type load scheduling unit load demand state grade
Elastic margin state rating
And zone elastic amplitude state level
Composition is carried out; supThe state space of a dispatching center, N is the number of regions;
setting the multi-type load containing type as M, if the area i does not consider the load j of a certain type, corresponding to the state number
Is 0; the total number of states N is obtained by the formula (19)up,s：
In the formula (19), NpeakThe maximum state grade of the real-time peak regulation demand of the power grid,
the photovoltaic output maximum state grade of the region i,
for the region iVRB energy storage cell charge-discharge maximum state level,
for the region i elasticity margin maximum state level,
for the state class with the largest elastic amplitude in the region i,
the maximum state grade of the load demand of the region i is set;
the dispatching center is arranged at decision time tkMaximum interval of lower random peak shaving demand power
The dispersion is 0 to Nap-1 to NapA plurality of levels, wherein,
for the scheduling centre at decision time tkLower total peak shaver power requirement, NapThe maximum discrete level is required for the total peak regulation of the dispatching center;
the peak-shaving task quantity distributed to the area i by the dispatching center is obtained by using the formula (20)
In the formula (20), the reaction mixture is,
for adjustingDegree center at decision time tkDescending the peak shaving task action allocated to the area i;
the peaker task action assignment constraint is established using equation (21):
in the formula (21), the compound represented by the formula,
Aia set of all possible peak shaver task motion vectors for the region i;
the formula (22) is used for obtaining the decision time t of the dispatching centerkAction vector of
In the formula (22), AupA set of all possible action vectors, namely an action set, for a scheduling center; the total number of actions of the dispatching center is Nup,a＝Nap；
Obtaining region i at decision time t using equation (23)kState of
In the formula (23), the compound represented by the formula,
is the state space of the region i;
the total number of states of the region i is obtained by equation (24)
Obtaining region i at decision time t using equation (25)kDownward movement
In the formula (25), the reaction mixture,
for region i at decision time tkThe lower VRB energy storage unit acts, and the three values are respectively a discharging action, an idle action and a charging action;
for region i at decision time tkThe lower load scheduling unit adjusting action comprises load reduction action capable of reducing load
Load shifting actions
For region i at decision time tkNext different actuation control actions;
is the set of all possible motion vectors in the region i, i.e. the motion set of the region i;
the total number of operations of the region i is obtained by the equation (26)
Step 5.2, defining the state transition process of the DTMDP model:
the state-of-charge transfer equation for the VRB energy storage unit is established using equation (27):
in the formula (27), N is the number of the single batteries connected in series by the electric pile, IdIn order to charge and discharge the current,
the total capacity of the VRB energy storage unit;
VRB energy storage unit of area i at current decision time tkThe state of charge of the battery,
taking charging and discharging actions for VRB energy storage unit
A later state of charge;
a reducible load state transition equation is established using equation (28):
in the formula (28), the reaction mixture is,
for region i at decision time tkTake a curtailment action
The latter can reduce the load demand situation,
for region i at decision time tkCan reduce the predicted power of the load demand,
decision time tkThe uncertain part of the load demand can be reduced,
the maximum discrete level of the load demand can be reduced;
the transferable load state transfer equation is established using equation (29):
in the formula (29), the reaction mixture,
for region i at decision time tkTake transfer action down
The latter transferable load demand situation,
at the end decision time t for region iK-1The next transfer action to be taken is,
for region i at decision time tkThe transferable load demand of (a) predicts the power,
as a decision time tkThe uncertain portion of the lower transferable load demand,
the maximum discrete level of transferable load demand;
step 5.3, establishing an objective function of the DTMDP model:
the upper layer cost in the decision period k is obtained by using the formula (30)
In the formula (30), ci,kGenerating a cost for the region i in the state transition process of the decision period k;
the starting and ending state of charge consistency constraint of the VRB energy storage unit is established by using the formula (31):
in the formula (31), the reaction mixture,
is the weight coefficient of the last state of the VRB energy storage unit,
and
respectively setting the actual capacity grade of the VRB energy storage unit at the last decision moment and the expected capacity grade;
step 5.4, establishing an optimization target of the DTMDP model:
obtaining scheduling center-in-strategy pi by using formula (32)upThe initial state is s0For a limited period of time to optimize the performance criterion
The upper layer optimization target is in a strategy set omegaupFind the optimal strategy
Obtaining region i in strategy pi using equation (33)dow,iThe initial state is s0For a limited period of time to optimize the performance criterion
The lower layer optimization target is in a strategy set omegadow,iFind the optimal strategy
Step 6, solving the DTMDP model established in the step 5 by adopting Q learning based on simulated annealing;
firstly, initializing parameters, learning parameters, upper and lower layer Q value tables, current learning step numbers and decision periods of a DTMDP model; then the upper and lower layers randomly select the corresponding action of the current state according to the strategy, generate the corresponding cost and update the Q value table; and repeatedly and iteratively updating the Q value table until the termination condition is met, and obtaining a scheduling strategy of each scheduling resource in each decision period meeting the peak regulation requirement of the scheduling center within one scheduling day.
Compared with the prior art, the invention has the beneficial effects that:
1. aiming at the scheduling optimization problem of a multi-region active power distribution system, the invention considers the peak regulation and the power demand elasticity of a power grid, utilizes various power elasticity resources to carry out cooperative optimization, and carries out strategy solution through a Q learning algorithm, thereby improving the load utilization rate and the peak regulation participation degree of each power consumption main body to a certain extent.
2. The invention adopts a layered learning algorithm, decomposes the active power distribution system into a plurality of different subsystems by copying boundary nodes, each subsystem is relatively independent, and performs overall coordination only by exchanging boundary node data, decomposes the original huge knowledge matrix into a plurality of smaller knowledge matrices, reduces the number of state action pairs, avoids dimension disaster to a certain extent, and is suitable for the problem of cooperative scheduling of a plurality of regions;
3. aiming at the scheduling problem of a multi-region active power distribution system, the simulated annealing algorithm is combined with Q learning, so that the exploration capability of the algorithm is further enhanced, the phenomenon of local optimization is avoided, and the most peak regulation strategy is more favorably obtained; meanwhile, the scheduling center peak shaving task allocation is carried out under the condition of considering the elastic amplitude of each region, so that the optimization time can be reduced to a certain extent; potential scheduling information of the active power distribution system can be further obtained by considering the elasticity of power demand, an intelligent solution is provided for scheduling of the active power distribution system, and stable and safe operation of the system is promoted.
Drawings
FIG. 1 is a schematic diagram of a multi-zone active power distribution system according to the present invention;
FIG. 2 is a flow chart of hierarchical coordinated scheduling as employed by the present invention;
fig. 3 is a flowchart of an algorithm for solving the problem of dynamic scheduling of a multi-zone active power distribution system according to the present invention.
Detailed Description
In this embodiment, a method for dynamically scheduling and optimizing a multi-region active power distribution system is applied to the multi-region active power distribution system shown in fig. 1, where the elastic power resources include: photovoltaic arrays, VRB energy storage devices and multi-type flexible loads in the region; in the process that the dispatching center distributes the peak shaving tasks to the areas, various electric power elastic resources and the day-ahead prediction of peak shaving requirements in each period are used as initial input, dispatching plan distribution is carried out under the condition that the elastic amplitude of each area is considered, and the dispatching plans are corrected in a rolling mode on a real-time scale.
Referring to fig. 2, the scheduling optimization method for the multi-zone active power distribution system is performed according to the following steps:
the area i includes: the system comprises an ith PV generating unit, an ith VRB energy storage unit and an ith load scheduling unit; the load types of the ith load scheduling unit include: the ith load-reducible load and the ith rigid load; wherein, the load type of load scheduling unit in the industrial park scheduling region still includes: the load can be transferred;
Step 3, establishing mathematical models of various load scheduling units, VRB energy storage units and PV power generation units in industrial, commercial, municipal and life parks, a region scheduling time attribute mathematical model and a peak shaving task allocation mechanism considering region elasticity amplitude:
step 3.1, establishing mathematical models of various types of loads of multiple areas under the elastic environment:
obtaining the minimum and maximum load reduction quantity constraints of the region i at the decision time t by using the formula (1):
in the formula (1), the reaction mixture is,
the maximum load amount of the reducible load of the area i at the decision time t is reduced;
actually reducing the load of the area i at the decision time t;
compared with the power environment without considering the elasticity of the power demand, the variable excitation control is set, so that the load can be reduced to generate a certain degree of upward deviation on the basis of the originally declared reduction upper limit.
The transferable load is characterized in that a user only needs to change the service time of electric energy without reducing the self life requirement or interrupting the production task, and the total electricity consumption of the user in a period is ensured to be unchanged, so that the increase and the decrease of the transferable load can be balanced. Considering that users have certain uncontrollable basic loads in the scheduling process, namely, the load increasing and decreasing amount in each time interval has certain limitation, the output constraint of transferable loads is obtained by using the formulas (2) to (4):
in the formulae (2) to (4),
and
respectively determining the maximum allowable load increase and the maximum allowable load decrease of the transferable load in the area i at the decision time t;
the load increment of the transferable load corresponding to the area i at the decision time t;
the load reduction quantity of the area i at the decision time t;
defining the accumulated increase and decrease load quantity after action is taken at each decision time from the initial time to the decision time t as
Suppose a single scheduling day has tK-1At the moment of decision, then
Is the remaining period tK-1T is an elastic margin of the transferable load, wherein the elastic margin can represent the fluctuation change of the transferable load range in the remaining time period in real time so as to influence the scheduling decision of the transferable load;
obtaining margin constraints of the region i at the decision time t by using the equations (5) and (6)
And transfer direction constraints
Thereby obtaining transferable load constraint of the area i at the decision time t
In the formulae (5) and (6), aacctrIs the elastic margin coefficient; e is a natural constant; alpha is alphadirSelecting a direction coefficient for the transfer motion;
the load amount is increased or decreased by the accumulated transfer to the t-1 moment;
when the load is increased and decreased correspondingly after the action is taken for the transferable load at the current decision time t and the load is set to be increased,
the value of (d) is positive, and when the load amount is reduced,
the value of (d) is negative, when not active,
is 0;
3.2, establishing a mathematical model of the output of the VRB energy storage unit:
in order to fully reflect the dynamic change characteristics of terminal voltage, terminal current, energy storage state of charge (SOC) and the like in the charging and discharging process of the all-vanadium redox flow battery (VRB), the constraint conditions of the VRB energy storage unit in one scheduling day are established by using an equation (7) to an equation (10), and the constraint conditions comprise the following steps: terminal voltage constraint, charge-discharge power constraint, charge state constraint and initial and final charge state consistency constraint:
in the formulae (7) to (10),
the upper limit and the lower limit of the end voltage of the VRB energy storage unit of the area i,
the minimum and maximum charge-discharge power of the VRB energy storage unit in the area i at the decision time t,
the actual charging and discharging power of the VRB energy storage unit at decision time t for region i,
the remaining capacity of the VRB energy storage cells for region i constrains the upper and lower limits,
for the remaining capacity of the VRB energy storage unit at decision time t for region i, ts、teTo schedule the beginning and end of the day, CconSetting the expected value of the state of charge of the VRB energy storage unit;
in actual operation, the SOC of the energy storage device is controlled to be 0.2-0.8 so as to ensure that the energy storage device works in a safety area, and the charging and discharging efficiency of the battery is improved.
3.3, establishing a mathematical model of photovoltaic power generation output:
obtaining a predicted value of the photovoltaic output power of the area i at the decision time t by using the formula (11)
In the formula (11), etapvThe photoelectric conversion efficiency; n ispvThe number of the photovoltaic cell panels; spvThe surface area of the photovoltaic cell panel receiving illumination is m2；
Solar radiation intensity in kW/m for area i at decision time t2；αpvThe order of magnitude is 10^ -3 in general for the temperature conversion coefficient,
outdoor temperature in units of area i at decision time t;
step 3.4, establishing a mathematical model of the regional scheduling time attribute:
when the cooperative scheduling of different areas is involved, the time dimension is introduced by considering the difference of different area characteristics. When a scheduling center issues peak shaving tasks and multi-elastic resource coordination adjustment, compared with the situation that longitudinal time difference is not considered, adjustment sequences or adjustment amount changes may exist in adjustment of different regional main bodies, and potential scheduling information of the active power distribution system is further mined. Considering that the scheduling time attribute is related to the self electricity utilization level and the electricity utilization characteristics of the region, the time attribute T of the region i at the decision time T is obtained by using an equation (12) and taking the absolute value of the difference between the electricity utilization load in each time period and the daily electricity utilization peak value of the region as an independent variable on the basis of the typical daily load curve of each regioni,t：
In the formula (12), the reaction mixture is,
is the time magnitude parameter of the region i; c is a constant, the magnitude of the order is visible
And in turn, the temperature of the molten metal is controlled,
when the value is MW, c can be 10^ 2; p is a radical ofi,tThe power output at decision time t for region i,
maximum output power of the region i on the dispatching day;
considering that transferable loads have a time strong constraint condition, the scheduling time attribute is temporarily limited in reducible loads.
Step 3.5, establishing a peak regulation task allocation mechanism considering the elastic amplitude of the region:
setting the elastic amplitude of the area i at the decision time t as Ei,tElastic amplitude Ei,tUpper bound of (2) which can reduce the upper bound of the load
+ transferable increased load upper bound
+ an energy storage discharge margin; elastic amplitude Ei,tLower bound of (2) reducible load lower bound + transferable reduced lower bound
+ an energy storage charging margin; the elastic amplitude of the region i is formed by the span between the upper and lower bounds;
compared with random task allocation, the method has the advantages that the peak regulation task allocation decision made by considering the elastic amplitude state information of each region at the current moment is taken into consideration when the peak regulation task is allocated, the system electric power elastic resource information is further utilized, and the power utilization efficiency is improved to a certain extent. And then, on the basis, performing rolling correction on the scheduling plan, and establishing an in-day real-time scheduling model considering the multi-period response characteristic of the demand response resource.
Step 4, modeling of continuous variable discretization and uncertain random variable dynamic change processes:
step 4.1, establishing a multi-region power grid peak regulation demand uncertainty model:
at decision time t, issuing the power grid to the maximum range interval of the random uncertain part of the peak regulation demand instruction of the area i in real time
Is dispersed into
In total
A plurality of levels, wherein,
for the maximum value of the upward fluctuation based on the peak shaver demand predicted power for the region i at the decision time t,
the maximum value of downward fluctuation based on the peak load demand prediction power of the area i at the decision time t;
maximum discrete levels of upward and downward fluctuation amounts based on the predicted power of the peak shaving demand of the region i;
and (3) obtaining the actual peak regulation demand of the power grid of the area i at the decision time t by using the formula (13):
in the formula (13), the reaction mixture is,
make decisions for region iThe peak shaving demand of the power grid at time t predicts power,
the power level of the uncertain part of the peak shaving demand of the power grid in the area i at the decision time t,
the minimum unit power of the power grid peak regulation instruction uncertainty part discretization of the area i at the decision time t is obtained;
step 4.2, establishing a photovoltaic output uncertain model;
the maximum range interval of the photovoltaic output uncertain part of the area i at the decision time t
Is dispersed into
In total
A plurality of levels, wherein,
the maximum value of the upward fluctuation based on the photovoltaic output predicted power of the area i at the decision time t,
the maximum value of the downward fluctuation based on the photovoltaic output predicted power of the area i at the decision time t;
maximum discrete levels of upward and downward fluctuation amounts based on the photovoltaic output predicted power of the region i;
obtaining the actual photovoltaic output of the region i at the decision time t by using the formula (14)
In the formula (14), the compound represented by the formula (I),
the power is predicted for the photovoltaic output at decision time t for zone i,
the power level of the uncertain photovoltaic output part of the area i at the decision time t is determined;
the minimum unit power of the area i after the photovoltaic output uncertain part is dispersed at decision time t is obtained;
step 4.3, establishing a multi-region uncertain model of each type of load demand:
the maximum range interval of the random uncertain part of the reducible load and the transferable load of the area i at the decision time t
And
respectively dispersed into corresponding state grades
And
in total
And
a plurality of levels, wherein,
and
respectively the maximum value of the upward fluctuation based on the forecasted power of the reducible load and the transferable load demand of the area i at the decision time t,
and
maximum values of downward fluctuation based on demand predicted power of reducible load and transferable load in the area i at decision time t respectively;
and
maximum discrete levels of upward and downward fluctuation amounts based on predicted power required by reducible load and transferable load in the region i respectively;
obtaining the actual power demand of the region i at the decision time t, which can reduce the load, by using the equations (15) to (16)
And the actual power demand of the load can be transferred
In the formulae (15) to (16),
power is predicted for the demand that region i can shed loads and transferable loads at decision time t,
respectively the power levels of the uncertain parts of the reducible load and the transferable load demand power in the area i at the decision time t;
respectively the minimum unit power of the region i after the uncertain part of the reducible load and the transferable load demand power is dispersed under the decision time t;
the process of the variable uncertain part changing along with the time is approximately described by a first-order Markov process, and the transition probability at each moment follows discrete Gaussian distribution taking the state of the variable uncertain part as the center.
Step 5, establishing a corresponding DTMDP model according to the peak regulation optimization problem of the multi-region active power distribution system considering the elasticity of power consumption requirements:
step 5.1, system state space and action set of the DTMDP model:
dividing a scheduling day into K e {0, 1.,. K-1}, and K decision periods; the time length of each decision period is delta t, and the decision time of the kth decision period is tkThe end time of the scheduling day is tK-1；
The formula (17) -formula (18) is used for obtaining the decision time t of the scheduling centerkState of
In the formulae (17) to (18),
as a decision time tkThe real-time peak regulation demand state grade of the lower power grid,
for region i at decision time tkEnvironmental information of the environment, photovoltaic output state class
VRB energy storage unit charging and discharging state grade
Multi-type load scheduling unit load demand state grade
Elastic margin state rating
And zone elastic amplitude state level
Composition is carried out; supThe state space of a dispatching center, N is the number of regions;
setting the multi-type load containing type as M, if the area i does not consider the load j of a certain type, corresponding to the state number
Is 0; the total number of states N is obtained by the formula (19)up,s：
In the formula (19), NpeakThe maximum state grade of the real-time peak regulation demand of the power grid,
is a regionThe maximum state class of photovoltaic output of the field i,
for the region iVRB energy storage cell charge-discharge maximum state level,
for the region i elasticity margin maximum state level,
for the state class with the largest elastic amplitude in the region i,
the maximum state grade of the load demand of the region i is set;
the dispatching center is arranged at decision time tkMaximum interval of lower random peak shaving demand power
The dispersion is 0 to Nap-1 to NapA plurality of levels, wherein,
for the scheduling centre at decision time tkLower total peak shaver power requirement, NapThe maximum discrete level is required for the total peak regulation of the dispatching center;
the peak-shaving task quantity distributed to the area i by the dispatching center is obtained by using the formula (20)
In the formula (20), the reaction mixture is,
for the scheduling centre at decision time tkDescending the peak shaving task action allocated to the area i;
the peaker task action assignment constraint is established using equation (21):
in the formula (21), the compound represented by the formula,
Aia set of all possible peak shaver task motion vectors for the region i;
the formula (22) is used for obtaining the decision time t of the dispatching centerkAction vector of
In the formula (22), AupA set of all possible action vectors, namely an action set, for a scheduling center; the total number of actions of the dispatching center is Nup,a＝Nap；
Obtaining region i at decision time t using equation (23)kState of
In the formula (23), the compound represented by the formula,
is the state space of the region i;
the total number of states of the region i is obtained by equation (24)
Obtaining region i at decision time t using equation (25)kDownward movement
In the formula (25), the reaction mixture,
for region i at decision time tkThe lower VRB energy storage unit acts, and the three values are respectively a discharging action, an idle action and a charging action;
for region i at decision time tkThe lower load scheduling unit adjusting action comprises load reduction action capable of reducing load
Load shifting actions
For region i at decision time tkNext different actuation control actions;
is the set of all possible motion vectors in the region i, i.e. the motion set of the region i;
the total number of operations of the region i is obtained by the equation (26)
Step 5.2, defining the state transition process of the DTMDP model:
the state-of-charge transfer equation for the VRB energy storage unit is established using equation (27):
in the formula (27), N is the number of the single batteries connected in series by the electric pile, IdIn order to charge and discharge the current,
the total capacity of the VRB energy storage unit;
VRB energy storage unit of area i at current decision time tkThe state of charge of the battery,
taking charging and discharging actions for VRB energy storage unit
A later state of charge;
a reducible load state transition equation is established using equation (28):
in the formula (28), the reaction mixture is,
for region i at decision time tkTake a curtailment action
Later can reduce the load demandIn the case of a situation in which,
for region i at decision time tkCan reduce the predicted power of the load demand,
decision time tkThe uncertain part of the load demand can be reduced,
the maximum discrete level of the load demand can be reduced;
considering that the transferable load needs to meet the constraint that the total amount is unchanged before and after the transfer, a state transfer equation of the transferable load is established by using an equation (29):
in the formula (29), the reaction mixture,
for region i at decision time tkTake transfer action down
The latter transferable load demand situation,
at the end decision time t for region iK-1The next transfer action to be taken is,
for region i at decision time tkThe transferable load demand of (a) predicts the power,
as a decision time tkThe uncertain portion of the lower transferable load demand,
the maximum discrete level of transferable load demand; the total amount before and after the transfer is not changed and restrained by the last moment adjusting action:
in the case of positive numbers, 0 and negative numbers,
the values of (a) correspond to-1, 0 and 1, respectively;
step 5.3, establishing an objective function of the DTMDP model:
the upper layer cost in the decision period k is obtained by using the formula (30)
Upper layer cost
Returning a cost c for each step of operation of each region of the lower layeri,kSuperposition of (2):
in the formula (30), ci,kGenerating a cost for the region i in the state transition process of the decision period k;
the active power distribution system scheduling is considered to be periodic, and the initial and final charge states of the energy storage device also meet the consistent constraint. The starting and ending state of charge consistency constraint of the VRB energy storage unit is established by using the formula (31):
in the formula (31), the reaction mixture,
is the weight coefficient of the last state of the VRB energy storage unit,
and
respectively setting the actual capacity grade of the VRB energy storage unit at the last decision moment and the expected capacity grade;
step 5.4, establishing an optimization target of the DTMDP model:
obtaining scheduling center-in-strategy pi by using formula (32)upThe initial state is s0For a limited period of time to optimize the performance criterion
The upper layer optimization target is in a strategy set omegaupFind the optimal strategy
Obtaining region i in strategy pi using equation (33)dow,iThe initial state is s0For a limited period of time to optimize the performance criterion
The lower layer optimization target is in a strategy set omegadow,iFind the optimal strategy
And 6, referring to fig. 3, solving the established DTMDP model by adopting simulated annealing-based Q learning according to the following steps:
and 6.1, initializing system model parameters. Comprising a single sample blockThe strategy cycle number K, the maximum level N of the task allocation of the dispatching centerpeakMaximum photovoltaic output rating in region i
Maximum grade of output force of energy storage device
Multi-type flexible load power adjustment maximum grade
Time of use electricity price
Coefficient of rebound load beta1、β2、β3Operation weight coefficient gammavrb、λpvEtc.;
and 6.2, initializing system learning parameters. Including the total number of sample tracks M, the learning rate of the dispatching center alphaupDiscount factor gammaupAnd learning rate update coefficient etaup(ii) a Learning rate of region i
Discount factor
And learning rate update coefficient
Simulated annealing temperature TtempAnd simulated annealing coefficient etatemp；
Step 6.3, initializing Q value table Q of dispatching centerupAnd Q value table of area i
Scheduling center and state data of each region to determine the state of the scheduling center
And initializing current learningThe step number m is equal to 0, and the current decision period k is equal to 0;
step 6.4 according to QupAnd greedy policy greedypolicyupSelecting the current state
Greedy action for down-per-region peak shaving task allocation
Simultaneously randomly selecting valid actions
If it is
The current scheduling center action is
Otherwise
Assigning peaking tasks to actions
Transmitting the status to each region, and observing the status of each region
Step 6.5, according to region i
And greedy strategy
Selecting a current state
Greedy action for lower corresponding region i
Simultaneously randomly selecting valid actions
If it is
The action of the current area i is
Otherwise
Observing the next period state of the dispatching center
And (3) counting the cost in the process and feeding the cost back to the dispatching center, and updating the Q value table of each area by using a formula (34):
step 6.6, update Q with equation (35)upIf K is less than K, skipping to step 6.4; otherwise, jumping to step 6.7:
step 6.7, executing the action selected by the current dispatching center and each area i
Calculating the cost generated in the process of executing action state transition in the decision period K
Updating the Q value table Q of the scheduling center and each area i by using the formula (36)up,
And (2) enabling m to be m +1:
step 6.8, if M is less than M, updating the learning rate alphaup:＝ηupαup，
And updates the temperature Ttemp:＝ηtempTtempReturning to the step 6.4; otherwise, finishing the learning optimization method to obtain a scheduling strategy of each scheduling resource in each decision period meeting the peak regulation requirement of the scheduling center within one scheduling day.
In conclusion, the invention can effectively deal with the randomness of each elastic resource in a multi-region active power distribution system and ensure the stable and safe operation of the system.
Claims (1)
1. A peak regulation optimization method of a multi-region active power distribution system considering power demand elasticity is characterized by comprising the following steps:
step 1, constructing a multi-region active power distribution system, comprising: the system comprises a dispatching center, an industrial park dispatching area, a commercial park dispatching area and municipal and residential life park dispatching areas; recording any one of the industrial park dispatching area, the commercial park dispatching area and the municipal and residential life park dispatching areas as an area i;
the area i comprises: the system comprises an ith PV generating unit, an ith VRB energy storage unit and an ith load scheduling unit; the load type of the ith load scheduling unit comprises: the ith load-reducible load and the ith rigid load; wherein, the load type of the load scheduling unit in the industrial park scheduling area further comprises: the load can be transferred;
step 2, determining predicted values of power grid peak regulation requirements, photovoltaic output and various types of load requirements based on historical data of the area i at any decision time t in a dispatching day; wherein, the peak load regulation demand of the power grid is recorded as
Photovoltaic output is noted
Reducible load demand among various types of load demands is noted
And transferable load demand is noted
Step 3, establishing mathematical models of various load scheduling units, VRB energy storage units and PV power generation units in industrial, commercial, municipal and life parks, a region scheduling time attribute mathematical model and a peak shaving task allocation mechanism considering region elasticity amplitude:
step 3.1, establishing mathematical models of various types of loads of multiple areas under the elastic environment:
obtaining the minimum and maximum load reduction quantity constraints of the region i at the decision time t by using the formula (1):
in the formula (1), the reaction mixture is,
the maximum load amount of the reducible load of the area i at the decision time t is reduced;
actually reducing the load of the area i at the decision time t;
the transferable load force constraint is obtained by using the formula (2) to the formula (4):
in the formulae (2) to (4),
and
respectively determining the maximum allowable load increase and the maximum allowable load decrease of the transferable load in the area i at the decision time t;
the load increment of the transferable load corresponding to the area i at the decision time t;
the load reduction quantity of the area i at the decision time t;
defining the accumulated increase and decrease load quantity after action is taken at each decision time from the initial time to the decision time t as
Suppose a single scheduling day has tK-1At the moment of decision, then
Is the remaining period tK-1-t is a margin of resilience to transferable load;
obtaining margin constraints of the region i at the decision time t by using the equations (5) and (6)
And transfer direction constraints
Thereby obtaining transferable load constraint of the area i at the decision time t
In the formulae (5) and (6), aacctrIs the elastic margin coefficient; e is a natural constant; alpha is alphadirSelecting a direction coefficient for the transfer motion;
the load amount is increased or decreased by the accumulated transfer to the t-1 moment;
when the load is increased and decreased correspondingly after the action is taken for the transferable load at the current decision time t and the load is set to be increased,
the value of (d) is positive, and when the load amount is reduced,
the value of (d) is negative, when not active,
is 0;
3.2, establishing a mathematical model of the output of the VRB energy storage unit:
establishing the constraint condition of the VRB energy storage unit in one scheduling day by using an equation (7) to an equation (10), wherein the constraint condition comprises the following steps: terminal voltage constraint, charge-discharge power constraint, charge state constraint and initial and final charge state consistency constraint:
in the formulae (7) to (10),
the upper limit and the lower limit of the end voltage of the VRB energy storage unit of the area i,
the minimum and maximum charge-discharge power of the VRB energy storage unit in the area i at the decision time t,
the actual charging and discharging power of the VRB energy storage unit at decision time t for region i,
the remaining capacity of the VRB energy storage cells for region i constrains the upper and lower limits,
for the remaining capacity of the VRB energy storage unit at decision time t for region i, ts、teTo schedule the beginning and end of the day, CconSetting the expected value of the state of charge of the VRB energy storage unit;
3.3, establishing a mathematical model of photovoltaic power generation output:
obtaining a predicted value of the photovoltaic output power of the area i at the decision time t by using the formula (11)
In the formula (11), etapvThe photoelectric conversion efficiency; n ispvThe number of the photovoltaic cell panels; spvThe surface area of the photovoltaic cell panel receiving illumination is increased;
the solar radiation intensity of the area i at the decision time t; alpha is alphapvIn order to be the temperature conversion coefficient,
the outdoor temperature of the area i at the decision time t;
step 3.4, establishing a mathematical model of the regional scheduling time attribute:
obtaining the time attribute T of the area i at the decision time T by using the formula (12)i,t：
In the formula (12), the reaction mixture is,
is the time magnitude parameter of the region i; c is a constant; p is a radical ofi,tThe power output at decision time t for region i,
maximum output power of the region i on the dispatching day;
step 3.5, establishing a peak regulation task allocation mechanism considering the elastic amplitude of the region:
setting the elastic amplitude of the area i at the decision time t as Ei,tElastic amplitude Ei,tUpper bound of (2) which can reduce the upper bound of the load
Transferable increased load upper bound
Energy storage and discharge allowance; elastic amplitude Ei,tLower bound of (2) reducible load lower bound + transferable reduced lower bound
Energy storage charging allowance; the elastic amplitude of the region i is formed by the span between the upper and lower bounds;
step 4, modeling of continuous variable discretization and uncertain random variable dynamic change processes:
step 4.1, establishing a multi-region power grid peak regulation demand uncertainty model:
at decision time t, issuing the power grid to the maximum range interval of the random uncertain part of the peak regulation demand instruction of the area i in real time
Is dispersed into
In total
A plurality of levels, wherein,
for the maximum value of the upward fluctuation based on the peak shaver demand predicted power for the region i at the decision time t,
the maximum value of downward fluctuation based on the peak load demand prediction power of the area i at the decision time t;
maximum discrete levels of upward and downward fluctuation amounts based on the predicted power of the peak shaving demand of the region i;
and (3) obtaining the actual peak regulation demand of the power grid of the area i at the decision time t by using the formula (13):
in the formula (13), the reaction mixture is,
predicting power for the peak shaver demand of the power grid in the area i at the decision time t,
the power level of the uncertain part of the peak shaving demand of the power grid in the area i at the decision time t,
the minimum unit power of the power grid peak regulation instruction uncertainty part discretization of the area i at the decision time t is obtained;
step 4.2, establishing a photovoltaic output uncertain model;
the maximum range interval of the photovoltaic output uncertain part of the area i at the decision time t
Is dispersed into
In total
A plurality of levels, wherein,
the maximum value of the upward fluctuation based on the photovoltaic output predicted power of the area i at the decision time t,
the maximum value of the downward fluctuation based on the photovoltaic output predicted power of the area i at the decision time t;
maximum discrete levels of upward and downward fluctuation amounts based on the photovoltaic output predicted power of the region i;
obtaining the actual photovoltaic output of the region i at the decision time t by using the formula (14)
In the formula (14), the compound represented by the formula (I),
the power is predicted for the photovoltaic output at decision time t for zone i,
for the indeterminate part of the photovoltaic output of zone i at decision time tA power level;
the minimum unit power of the area i after the photovoltaic output uncertain part is dispersed at decision time t is obtained;
step 4.3, establishing a multi-region uncertain model of each type of load demand:
the maximum range interval of the random uncertain part of the reducible load and the transferable load of the area i at the decision time t
And
respectively dispersed into corresponding state grades
And
in total
And
a plurality of levels, wherein,
and
respectively the maximum value of the upward fluctuation based on the forecasted power of the reducible load and the transferable load demand of the area i at the decision time t,
and
maximum values of downward fluctuation based on demand predicted power of reducible load and transferable load in the area i at decision time t respectively;
and
maximum discrete levels of upward and downward fluctuation amounts based on predicted power required by reducible load and transferable load in the region i respectively;
obtaining the actual power demand of the region i at the decision time t, which can reduce the load, by using the equations (15) to (16)
And the actual power demand of the load can be transferred
In the formulae (15) to (16),
power is predicted for the demand that region i can shed loads and transferable loads at decision time t,
the uncertain part of the required power of the reducible load and the transferable load in the area i at the decision time tA fractional power level;
respectively the minimum unit power of the region i after the uncertain part of the reducible load and the transferable load demand power is dispersed under the decision time t;
step 5, establishing a corresponding DTMDP model according to the peak regulation optimization problem of the multi-region active power distribution system considering the elasticity of power consumption requirements:
step 5.1, system state space and action set of the DTMDP model:
dividing a scheduling day into K e {0, 1.,. K-1}, and K decision periods; the time length of each decision period is delta t, and the decision time of the kth decision period is tkThe end time of the scheduling day is tK-1；
The formula (17) -formula (18) is used for obtaining the decision time t of the scheduling centerkState of
In the formulae (17) to (18),
as a decision time tkThe real-time peak regulation demand state grade of the lower power grid,
for region i at decision time tkEnvironmental information of the environment, photovoltaic output state class
VRB energy storage unit charging and discharging state grade
Multi-type load scheduling unit load demand state grade
Elastic margin state rating
And zone elastic amplitude state level
Composition is carried out; supThe state space of a dispatching center, N is the number of regions;
setting the multi-type load containing type as M, if the area i does not consider the load j of a certain type, corresponding to the state number
Is 0; the total number of states N is obtained by the formula (19)up,s：
In the formula (19), NpeakThe maximum state grade of the real-time peak regulation demand of the power grid,
the photovoltaic output maximum state grade of the region i,
for the region iVRB energy storage cell charge-discharge maximum state level,
is a regioni the maximum state level of the margin of elasticity,
for the state class with the largest elastic amplitude in the region i,
the maximum state grade of the load demand of the region i is set;
the dispatching center is arranged at decision time tkMaximum interval of lower random peak shaving demand power
The dispersion is 0 to Nap-1 to NapA plurality of levels, wherein,
for the scheduling centre at decision time tkLower total peak shaver power requirement, NapThe maximum discrete level is required for the total peak regulation of the dispatching center;
the peak-shaving task quantity distributed to the area i by the dispatching center is obtained by using the formula (20)
In the formula (20), the reaction mixture is,
for the scheduling centre at decision time tkDescending the peak shaving task action allocated to the area i;
the peaker task action assignment constraint is established using equation (21):
in the formula (21), the compound represented by the formula,
Aia set of all possible peak shaver task motion vectors for the region i;
the formula (22) is used for obtaining the decision time t of the dispatching centerkAction vector of
In the formula (22), AupA set of all possible action vectors, namely an action set, for a scheduling center; the total number of actions of the dispatching center is Nup,a＝Nap；
Obtaining region i at decision time t using equation (23)kState of
In the formula (23), the compound represented by the formula,
is the state space of the region i;
the total number of states of the region i is obtained by equation (24)
Obtaining region i at decision time t using equation (25)kDownward movement
In the formula (25), the reaction mixture,
for region i at decision time tkThe lower VRB energy storage unit acts, and the three values are respectively a discharging action, an idle action and a charging action;
for region i at decision time tkThe lower load scheduling unit adjusting action comprises load reduction action capable of reducing load
Load shifting actions
For region i at decision time tkNext different actuation control actions;
is the set of all possible motion vectors in the region i, i.e. the motion set of the region i;
the total number of operations of the region i is obtained by the equation (26)
Step 5.2, defining the state transition process of the DTMDP model:
the state-of-charge transfer equation for the VRB energy storage unit is established using equation (27):
in the formula (27), N is the number of the single batteries connected in series by the electric pile, IdIn order to charge and discharge the current,
the total capacity of the VRB energy storage unit;
VRB energy storage unit of area i at current decision time tkThe state of charge of the battery,
taking charging and discharging actions for VRB energy storage unit
A later state of charge;
a reducible load state transition equation is established using equation (28):
in the formula (28), the reaction mixture is,
for region i at decision time tkTake a curtailment action
The latter can reduce the load demand situation,
for region i at decision time tkCan reduce the predicted power of the load demand,
decision time tkThe uncertain part of the load demand can be reduced,
the maximum discrete level of the load demand can be reduced;
the transferable load state transfer equation is established using equation (29):
in the formula (29), the reaction mixture,
for region i at decision time tkTake transfer action down
The latter transferable load demand situation,
at the end decision time t for region iK-1The next transfer action to be taken is,
for region i at decision time tkThe transferable load demand of (a) predicts the power,
as a decision time tkThe uncertain portion of the lower transferable load demand,
the maximum discrete level of transferable load demand;
step 5.3, establishing an objective function of the DTMDP model:
the upper layer cost in the decision period k is obtained by using the formula (30)
In the formula (30), ci,kGenerating a cost for the region i in the state transition process of the decision period k;
the starting and ending state of charge consistency constraint of the VRB energy storage unit is established by using the formula (31):
in the formula (31), the reaction mixture,
is the weight coefficient of the last state of the VRB energy storage unit,
and
respectively setting the actual capacity grade of the VRB energy storage unit at the last decision moment and the expected capacity grade;
step 5.4, establishing an optimization target of the DTMDP model:
obtaining scheduling center-in-strategy pi by using formula (32)upThe initial state is s0Is optimized within a limited period of timePerformance criteria
The upper layer optimization target is in a strategy set omegaupFind the optimal strategy
Obtaining region i in strategy pi using equation (33)dow,iThe initial state is s0For a limited period of time to optimize the performance criterion
The lower layer optimization target is in a strategy set omegadow,iFind the optimal strategy
Step 6, solving the DTMDP model established in the step 5 by adopting Q learning based on simulated annealing;
firstly, initializing parameters, learning parameters, upper and lower layer Q value tables, current learning step numbers and decision periods of a DTMDP model; then the upper and lower layers randomly select the corresponding action of the current state according to the strategy, generate the corresponding cost and update the Q value table; and repeatedly and iteratively updating the Q value table until the termination condition is met, and obtaining a scheduling strategy of each scheduling resource in each decision period meeting the peak regulation requirement of the scheduling center within one scheduling day.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110366783.5A CN112952847B (en) | 2021-04-06 | 2021-04-06 | Multi-region active power distribution system peak regulation optimization method considering electricity demand elasticity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110366783.5A CN112952847B (en) | 2021-04-06 | 2021-04-06 | Multi-region active power distribution system peak regulation optimization method considering electricity demand elasticity |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112952847A true CN112952847A (en) | 2021-06-11 |
| CN112952847B CN112952847B (en) | 2022-09-16 |
Family
ID=76232390
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110366783.5A Active CN112952847B (en) | 2021-04-06 | 2021-04-06 | Multi-region active power distribution system peak regulation optimization method considering electricity demand elasticity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112952847B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113690878A (en) * | 2021-08-03 | 2021-11-23 | 北京京能能源技术研究有限责任公司 | Micro-grid three-switch control method |
| CN113706018A (en) * | 2021-08-27 | 2021-11-26 | 广东电网有限责任公司 | User peak regulation behavior evaluation model establishing method, evaluation method and device |
| CN116760043A (en) * | 2023-05-29 | 2023-09-15 | 中国建筑科学研究院有限公司 | Heat pump system power grid peak shaving potential and effect evaluation method considering multi-dimension index |
| CN117937474A (en) * | 2024-03-20 | 2024-04-26 | 保定博堃元信息科技有限公司 | New energy station energy storage management method and system |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160189175A1 (en) * | 2014-12-24 | 2016-06-30 | Hanshuang LI | System and method of sensitivity-driven pricing service for non-stationary demand management |
| CN105741017A (en) * | 2016-01-22 | 2016-07-06 | 江苏省电力公司电力科学研究院 | User demand response assessment method in intelligent power grid environment |
| CN106777487A (en) * | 2016-11-18 | 2017-05-31 | 清华大学 | A kind of credible capacity calculation methods of the photovoltaic plant containing energy-storage system and system |
| CN108964042A (en) * | 2018-07-24 | 2018-12-07 | 合肥工业大学 | Regional power grid operating point method for optimizing scheduling based on depth Q network |
| CN109103912A (en) * | 2018-07-18 | 2018-12-28 | 合肥工业大学 | Consider the industrial park active distribution system method for optimizing scheduling of peaking demand of power grid |
| WO2019035527A1 (en) * | 2017-08-17 | 2019-02-21 | 한국전력공사 | Blockchain-based power trading operation system, method therefor, and computer readable storage medium that stores said method |
| CN109494727A (en) * | 2018-11-30 | 2019-03-19 | 国网江西省电力有限公司经济技术研究院 | Consider the active and idle coordination optimization operation method of power distribution network of demand response |
| CN109524958A (en) * | 2018-11-08 | 2019-03-26 | 国网浙江省电力有限公司经济技术研究院 | Consider the electric system flexibility Optimization Scheduling of depth peak regulation and demand response |
| CN110112767A (en) * | 2019-03-19 | 2019-08-09 | 华北电力大学 | The polymorphic Demand-side load of wide area participates in the lotus source optimization control method of peak-load regulating |
| CN111628503A (en) * | 2020-06-20 | 2020-09-04 | 东北电力大学 | Day-ahead-day two-stage rolling optimization scheduling method considering generalized energy storage and thermal power combined peak shaving |
-
2021
- 2021-04-06 CN CN202110366783.5A patent/CN112952847B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160189175A1 (en) * | 2014-12-24 | 2016-06-30 | Hanshuang LI | System and method of sensitivity-driven pricing service for non-stationary demand management |
| CN105741017A (en) * | 2016-01-22 | 2016-07-06 | 江苏省电力公司电力科学研究院 | User demand response assessment method in intelligent power grid environment |
| CN106777487A (en) * | 2016-11-18 | 2017-05-31 | 清华大学 | A kind of credible capacity calculation methods of the photovoltaic plant containing energy-storage system and system |
| WO2019035527A1 (en) * | 2017-08-17 | 2019-02-21 | 한국전력공사 | Blockchain-based power trading operation system, method therefor, and computer readable storage medium that stores said method |
| CN109103912A (en) * | 2018-07-18 | 2018-12-28 | 合肥工业大学 | Consider the industrial park active distribution system method for optimizing scheduling of peaking demand of power grid |
| CN108964042A (en) * | 2018-07-24 | 2018-12-07 | 合肥工业大学 | Regional power grid operating point method for optimizing scheduling based on depth Q network |
| CN109524958A (en) * | 2018-11-08 | 2019-03-26 | 国网浙江省电力有限公司经济技术研究院 | Consider the electric system flexibility Optimization Scheduling of depth peak regulation and demand response |
| CN109494727A (en) * | 2018-11-30 | 2019-03-19 | 国网江西省电力有限公司经济技术研究院 | Consider the active and idle coordination optimization operation method of power distribution network of demand response |
| CN110112767A (en) * | 2019-03-19 | 2019-08-09 | 华北电力大学 | The polymorphic Demand-side load of wide area participates in the lotus source optimization control method of peak-load regulating |
| CN111628503A (en) * | 2020-06-20 | 2020-09-04 | 东北电力大学 | Day-ahead-day two-stage rolling optimization scheduling method considering generalized energy storage and thermal power combined peak shaving |
Non-Patent Citations (4)
| Title |
|---|
| BERK CELIK ET AL.: "Quantifying the Impact of Solar Photovoltaic and Energy Storage Assets on the Performance of a Residential Energy Aggregator", 《IEEE TRANSACTIONS ON SUSTAINABLE ENERGY》, vol. 11, no. 1, 31 January 2020 (2020-01-31), pages 405 - 414, XP011762385, DOI: 10.1109/TSTE.2019.2892603 * |
| XIAO-YAN GE ET AL.: "Energy-Sustainable Traffic Signal Timings for a Congested Road Network With Heterogeneous Users", 《 IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS》, vol. 15, no. 3, 30 June 2014 (2014-06-30), pages 1016 - 1025, XP011549671, DOI: 10.1109/TITS.2013.2291612 * |
| 吴熙等: "基于Q学习算法的综合能源系统韧性提升方法", 《电力自动化设备》, vol. 40, no. 04, 30 April 2020 (2020-04-30), pages 146 - 152 * |
| 曹筱欧: "适应间歇式新能源发电负荷调峰策略研究", 《微型电脑应用》, vol. 35, no. 1, 31 December 2019 (2019-12-31), pages 91 - 97 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113690878A (en) * | 2021-08-03 | 2021-11-23 | 北京京能能源技术研究有限责任公司 | Micro-grid three-switch control method |
| CN113690878B (en) * | 2021-08-03 | 2023-11-21 | 北京京能能源技术研究有限责任公司 | Three-phase switching control method for micro-grid |
| CN113706018A (en) * | 2021-08-27 | 2021-11-26 | 广东电网有限责任公司 | User peak regulation behavior evaluation model establishing method, evaluation method and device |
| CN116760043A (en) * | 2023-05-29 | 2023-09-15 | 中国建筑科学研究院有限公司 | Heat pump system power grid peak shaving potential and effect evaluation method considering multi-dimension index |
| CN117937474A (en) * | 2024-03-20 | 2024-04-26 | 保定博堃元信息科技有限公司 | New energy station energy storage management method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112952847B (en) | 2022-09-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112952847B (en) | Multi-region active power distribution system peak regulation optimization method considering electricity demand elasticity | |
| Hossain et al. | Modified PSO algorithm for real-time energy management in grid-connected microgrids | |
| CN109103912B (en) | Industrial park active power distribution system scheduling optimization method considering power grid peak regulation requirements | |
| CN106532751B (en) | A kind of distributed generation resource efficiency optimization method and system | |
| CN111293682B (en) | Multi-microgrid energy management method based on cooperative model predictive control | |
| CN104158203A (en) | Micro-grid power supply capacity optimization configuration method | |
| CN112800658A (en) | Active power distribution network scheduling method considering source storage load interaction | |
| CN109114662B (en) | Heating control method and system of electrothermal energy storage device based on multiple intelligent agents | |
| CN112508325B (en) | Household micro-grid multi-time scale electric energy scheduling method | |
| CN113765154B (en) | High-elasticity power grid source network load storage multielement collaborative optimization control method | |
| CN114243794B (en) | Wind, light and water multi-energy complementary system flexibility requirement quantification and coordination optimization method | |
| CN115147245B (en) | Virtual power plant optimal scheduling method for industrial load participating in peak shaving auxiliary service | |
| CN117060470B (en) | Power distribution network voltage optimization control method based on flexible resources | |
| Dong et al. | Optimal scheduling framework of electricity-gas-heat integrated energy system based on asynchronous advantage actor-critic algorithm | |
| CN115036914A (en) | Power grid energy storage double-layer optimization method and system considering flexibility and new energy consumption | |
| CN111817296A (en) | Electric energy scheduling method and system for micro-grid | |
| CN111709574A (en) | Distributed cluster configuration scheduling method, computer equipment and storage medium | |
| CN114462854A (en) | Hierarchical scheduling method and system containing new energy and electric vehicle grid connection | |
| Tang et al. | Schedule of air-conditioning systems with thermal energy storage considering wind power forecast errors | |
| CN117559526A (en) | Router-simulated energy regulation and control method based on optical storage and charging integrated charging station | |
| Desdouits et al. | Multisource elevator energy optimization and control | |
| Yunjia et al. | Independent microgrid day-ahead optimization based on demand response | |
| CN116961008A (en) | Micro-grid capacity double-layer optimization method considering power spring and load demand response | |
| CN116384655A (en) | New energy consumption-oriented source network load side energy storage system optimization planning method and system | |
| CN114421538A (en) | Source-storage collaborative planning configuration method containing source-load time sequence similarity constraint |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220804 Address after: 230001 No. 193, Tunxi Road, Hefei, Anhui Applicant after: Hefei University of Technology Applicant after: State Grid Anhui Electric Power Company Address before: Tunxi road in Baohe District of Hefei city of Anhui Province, No. 193 230009 Applicant before: Hefei University of Technology |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |