CN116663820A - Comprehensive energy system energy management method under demand response - Google Patents
Comprehensive energy system energy management method under demand response Download PDFInfo
- Publication number
- CN116663820A CN116663820A CN202310588505.3A CN202310588505A CN116663820A CN 116663820 A CN116663820 A CN 116663820A CN 202310588505 A CN202310588505 A CN 202310588505A CN 116663820 A CN116663820 A CN 116663820A
- Authority
- CN
- China
- Prior art keywords
- energy
- power
- network
- decision time
- cold
- 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
Links
- 230000004044 response Effects 0.000 title claims abstract description 123
- 238000007726 management method Methods 0.000 title claims abstract description 30
- 238000005457 optimization Methods 0.000 claims abstract description 17
- 238000013528 artificial neural network Methods 0.000 claims abstract description 16
- 238000004519 manufacturing process Methods 0.000 claims abstract description 9
- 230000003993 interaction Effects 0.000 claims abstract description 6
- 238000000034 method Methods 0.000 claims description 83
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 claims description 78
- 239000007789 gas Substances 0.000 claims description 54
- 230000008569 process Effects 0.000 claims description 45
- 239000003345 natural gas Substances 0.000 claims description 39
- 238000005338 heat storage Methods 0.000 claims description 36
- 230000006870 function Effects 0.000 claims description 28
- 238000003860 storage Methods 0.000 claims description 28
- 238000010248 power generation Methods 0.000 claims description 24
- 230000008901 benefit Effects 0.000 claims description 19
- 239000002918 waste heat Substances 0.000 claims description 18
- 230000009471 action Effects 0.000 claims description 17
- 238000006243 chemical reaction Methods 0.000 claims description 17
- 238000009826 distribution Methods 0.000 claims description 17
- 238000004146 energy storage Methods 0.000 claims description 17
- 238000010521 absorption reaction Methods 0.000 claims description 15
- 238000004422 calculation algorithm Methods 0.000 claims description 15
- 238000001816 cooling Methods 0.000 claims description 15
- 230000000875 corresponding effect Effects 0.000 claims description 15
- 238000012423 maintenance Methods 0.000 claims description 15
- 230000005611 electricity Effects 0.000 claims description 13
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 claims description 12
- 229910052799 carbon Inorganic materials 0.000 claims description 12
- 238000005485 electric heating Methods 0.000 claims description 12
- 238000004378 air conditioning Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 7
- 230000007613 environmental effect Effects 0.000 claims description 6
- 238000010438 heat treatment Methods 0.000 claims description 6
- 238000005286 illumination Methods 0.000 claims description 6
- 230000009467 reduction Effects 0.000 claims description 6
- 238000005057 refrigeration Methods 0.000 claims description 6
- 230000006399 behavior Effects 0.000 claims description 4
- 230000010354 integration Effects 0.000 claims description 4
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 3
- 241000084490 Esenbeckia delta Species 0.000 claims description 3
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000001276 controlling effect Effects 0.000 claims description 3
- 230000005284 excitation Effects 0.000 claims description 3
- 239000003546 flue gas Substances 0.000 claims description 3
- XGFJCRNRWOXGQM-UHFFFAOYSA-N hot-2 Chemical compound CCSC1=CC(OC)=C(CCNO)C=C1OC XGFJCRNRWOXGQM-UHFFFAOYSA-N 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000011084 recovery Methods 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 238000004088 simulation Methods 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 238000005265 energy consumption Methods 0.000 abstract 1
- 230000002787 reinforcement Effects 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000009977 dual effect Effects 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 241000135164 Timea Species 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
-
- 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
Abstract
The invention belongs to the technical field of energy management, and provides an energy management method of a comprehensive energy system under demand response, which comprises the following steps: step 1: building an energy equipment model of a comprehensive energy system, an energy consumption behavior model of a user and an operation optimization model of the user; step 2: constructing a neural network required by TD 3; step 3: according to the model established in the step 1, the energy management center is utilized to interact with the comprehensive energy system, history information obtained by interaction is put into an experience pool, and meanwhile, the TD3 is utilized to realize optimization of a strategy network and a value network; step 4: and (3) controlling the comprehensive energy system according to the optimized strategy network obtained in the step (3) so as to participate in the power demand response. The invention is beneficial to reasonably arranging the production plan of the internal energy equipment and the energy utilization plan of the internal users of the comprehensive energy system, further improves the demand response potential of the system, and maintains the safe, stable and economic operation of the comprehensive energy system while completing the demand response plan of the power grid.
Description
Technical Field
The invention belongs to the technical field of energy management, and particularly relates to an energy management method of an integrated energy system under demand response.
Background
The environmental pollution and energy shortage problems are increasingly serious in the world today, and the improvement of the energy utilization efficiency is urgently needed. The comprehensive energy system takes multi-energy coupling production as a basis and takes multi-energy network cooperation as a dispatching core, breaks through the existing mode of separate planning and independent operation of each energy source, and can realize the cooperation, optimization and mutual aid of the multi-energy sources. The power demand response is an important mode for implementing power demand management, and mainly guides a user to adjust own power consumption behavior through modes such as electric energy price, policy incentive and the like, so that the phenomenon of power supply shortage can be effectively relieved, and the safety and stability of power grid operation are improved. With the gradual implementation of provincial power demand policies of Guangdong, anhui and the like, the comprehensive energy system can also participate in power demand response, can change the energy demand on a power grid by adjusting own energy production plans and the like, and has great demand response adjustment potential.
Solving the optimal scheduling problem of the comprehensive energy system mainly comprises three solving ideas of nonlinear method solving, intelligent algorithm solving and linear simplified solving. In the former two methods, in the solution of the high-dimensional nonlinear problem with more and more intimate coupling of electricity, heat and cold, the operation time is too long, the requirement of on-line calculation is difficult to meet, and the linearization treatment is easy to calculate and solve, but the condition of larger error of the solution result is unavoidable. With the rise and development of artificial intelligence technology, deep reinforcement learning is increasingly emphasized in the optimization control of integrated energy systems. The deep reinforcement learning has strong autonomous learning capability, can acquire history experience from massive history data, can take different actions under different system states, learn knowledge from rewards and return to acquire an optimal strategy, can not depend on detailed and accurate model information in the whole process of interaction with the environment, is easier to realize in an actual scheduling scene compared with the traditional method, and can realize real-time operation decision of the system by utilizing an offline learning and online decision making mode.
Disclosure of Invention
Aiming at the problems in the prior art of energy management of the existing comprehensive energy system, the invention provides an energy management method of the comprehensive energy system under the requirement response, which can realize safe, stable and economic operation of the comprehensive energy system under the influence of a plurality of random factors when the comprehensive energy system participates in the power requirement response.
In order to achieve the above purpose, the invention adopts the following technical scheme:
an integrated energy system energy management method under demand response comprises an integrated energy system, wherein the integrated energy system comprises an energy management center (EMS), energy conversion equipment, energy storage equipment and multiple types of users, the energy conversion equipment comprises a Gas Turbine (GT), a waste Heat Boiler (HB), a Gas Boiler (GB), an Absorption Refrigerator (AR), an electric heat pump (EB), an electric refrigerating unit (ER) and a photovoltaic power generation device (PV), the energy storage equipment comprises a storage battery (ES) and a heat storage tank (HS), the multiple types of users comprise public service users and commercial users, the loads of the users are three energy forms of electric loads, heat loads and cold loads, the energy management center is an integration center of energy and information, the energy management center is used for externally receiving information of an electric power supplier and a natural gas supplier, the information of electricity price, the natural gas price and a demand response plan issued by an electric company can be obtained, the purchase amount of electric energy and natural gas energy can be controlled, inherent response offers can be issued to the multiple types of users to adjust the user's use behaviors and control the energy equipment in the system, the energy management system can be controlled, the safety response is completed by the system can be completed by the specific energy management system, the energy management center is safe and the method is characterized by the following the economical response operation steps of the system, and the energy management system can be completed simultaneously:
Step 1: the integrated energy system participates in the power demand response by taking one day as a scheduling period, and the total aim of the optimized operation is to make the total profit of the system operation for one day maximum on the basis of completing the power demand response aim, so that an operation optimization model of the integrated energy system participating in the power demand response needs to be established:
step 1.1, establishing a power demand response plan, photovoltaic output randomness, outdoor temperature randomness and a user electric heating and cooling load randomness model;
step 1.2, establishing an energy conversion device and energy storage device scheduling model of the comprehensive energy system;
step 1.3, establishing a multi-type user response characteristic model in the comprehensive energy system;
step 1.4, establishing a dispatching optimization model of the comprehensive energy system participating in power demand response;
step 2: deep neural networks required to build the dual delay depth deterministic strategy gradient algorithm (Twin Delayed Deep Deterministic Policy Gradient Algorithm, TD 3):
the TD3 is provided with three independent neural networks, namely a strategy network, a value network 1 and a value network 2, each network is provided with a respective target network, namely a target strategy network, a target value network 1 and a target value network 2, and the six neural networks are all fully connected neural networks and comprise an input layer, a hidden layer and an output layer, wherein the strategy network and the target strategy network have the same structure, and the value network 1, the value network 2, the target value network 1 and the target value network 2 have the same structure;
Step 3: according to the model established in the step 1, the energy management center is utilized to interact with the comprehensive energy system, history interaction information is obtained and stored in an experience pool, and the TD3 is utilized to realize iterative updating and optimization of the strategy network, the value network 1 and the value network 2;
step 4: and 3, controlling the comprehensive energy system according to the strategy network obtained in the step 3:
by utilizing the feasibility of the established comprehensive energy system verification algorithm, at any decision time of system operation, information such as decision time, offer response quantity, operation state of the comprehensive energy system and the like is normalized and then is input into a strategy network of TD3, action information output by the strategy network is reversely normalized after forward propagation, and then actions which can be taken by the system at the current decision time can be obtained, and the situation that the system power demand response is completed after the comprehensive energy system repeatedly executes the operation in a scheduling period and the obtained selling energy benefit, response income, energy purchasing cost, excitation cost, maintenance cost and carbon tax cost are observed.
Further, the comprehensive energy system energy management method under demand response is characterized in that the power demand response plan, the photovoltaic output randomness, the outdoor temperature randomness and the user electric heating and cooling load randomness model in step 1.1 comprise the following specific steps:
Determining an amount of response to participate in the power demand response:
determining each decision time t of comprehensive energy system in one scheduling period k Response amount P 'of the power demand response of (a)' peak (t k ) Obtaining a demand response target of the comprehensive energy system at each decision moment in a scheduling period:
{P′ peak (t 1 ),P′ peak (t 1 ),...,P′ peak (t k ),...,P′ peak (t K )}
photovoltaic output randomness:
error Δp of photovoltaic output pv Mean value of compliance mu pv Variance is sigma pv Normal distribution of (a), i.e. ΔP pv ～N(μ pv ,σ pv 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the photovoltaic output is P pv,0 (t k ) The actual value P of the photovoltaic output can be obtained after the random error is superimposed pv (t k ) The method comprises the following steps:
P pv (t k )＝P pv,0 (t k )+ΔP pv
temperature randomness:
error delta T of temperature temp Mean value of compliance mu temp Variance is sigma temp Normal distribution of (a), i.e. delta T temp ～N(μ temp ,σ temp 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the temperature is T temp,0 (t k ) The actual value T of the temperature can be obtained after the random error is superimposed temp (t k ) The method comprises the following steps:
T temp (t k )＝T temp,0 (t k )+ΔT temp
electrical load randomness:
error Δl of electrical load ele Mean value of compliance mu ele Variance is sigma ele Normal distribution of (L), i.e. DeltaL ele ～N(μ ele ,σ ele 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the electric load is L ele,0 (t k ) The actual value L of the electric load can be obtained after the random error is superimposed ele (t k ) The method comprises the following steps:
L ele (t k )＝L ele,0 (t k )+ΔL ele
heat load randomness:
error Δl of thermal load hot Mean value of compliance mu hot Variance is sigma hot Normal distribution of (L), i.e. DeltaL hot ～N(μ hot ,σ hot 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the heat load is L hot,0 (t k ) The actual value L of the heat load can be obtained after the random errors are superimposed hot (t k ) The method comprises the following steps:
L hot (t k )＝L hot,0 (t k )+ΔL hot
cold load randomness:
error Δl of cooling load cold Mean value of compliance mu cold Variance is sigma cold Normal distribution of (L), i.e. DeltaL cold ～N(μ cold ,σ cold 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the cooling load is L cold,0 (t k ) The actual value L of the cold load can be obtained after the random errors are superimposed cold (t k ) The method comprises the following steps:
L cold (t k )＝L cold,0 (t k )+ΔL cold
further, the energy management method of the integrated energy system under the demand response is characterized in that the energy conversion equipment and the energy storage equipment scheduling model of the integrated energy system in the step 1.2 specifically comprises the following steps:
gas turbine scheduling process:
gas turbines provide electrical and thermal energy by consuming natural gas, decision time t k The power output from the gas turbine is calculated as follows:
P gt (t k )＝V gt (t k )H ng η gt
wherein ,Pgt (t k) and Vgt (t k ) For the decision time t k Electric power output by the gas turbine and consumed natural gas; h ng Is the heat value of natural gas; η (eta) gt Is the power generation efficiency of the gas turbine.
The ratio of the output thermal power to the electric power of the gas turbine is a thermoelectric ratio b, which can be expressed as:
wherein ,Qgt (t k ) For the decision time t k Heating power of the gas turbine.
The waste heat boiler scheduling process comprises the following steps:
the waste heat boiler can improve the utilization efficiency of energy, and provides heat energy for users by absorbing waste heat in the high-temperature flue gas discharged by the gas turbine, and the mathematical model is as follows:
Q hb (t k )＝Q hb,0 (t k )η hb
wherein ,Qhb (t k) and Qhb,0 (t k ) Decision time t k The heat power absorbed and output by the waste heat boiler; η (eta) hb Is the waste heat recovery efficiency.
The gas boiler scheduling process comprises the following steps:
the gas-fired boiler consumes natural gas to provide heat energy, and the gas-fired boiler makes a decision at time t k The output thermal power is calculated as follows:
Q gb (t k )＝V gb (t k )H ng η gb
wherein ,Qgb (t k) and Vgb (t k ) Is thatDecision time t k The heat power output by the gas boiler and the consumed natural gas; η (eta) gb Is the heat generating efficiency of the gas boiler.
Scheduling process of absorption refrigerator:
the absorption refrigerator provides cold power by absorbing hot power, and at decision time t k The output cold power is calculated as follows:
H ar (t k )＝Q ar (t k )η ar
wherein ,Har (t k) and Qar (t k ) For time t k The cold power and the absorbed hot power output by the absorption refrigerator; η (eta) ar Is the refrigerating efficiency of the absorption refrigerator.
And (3) an electric heating pump scheduling process:
the electric heat pump converts low-grade heat energy into high-grade heat energy by consuming electric energy, and at decision time t k The thermal power output by the electric heat pump is calculated as follows:
Q eb (t k )＝P eb (t k )η eb
wherein ,Qeb (t k) and Peb (t k ) For the decision time t k The heat power output by the electric heat pump and the consumed electric power; η (eta) eb The heating efficiency of the electric heat pump is achieved.
Scheduling process of electric refrigerating unit:
the electric refrigerating unit provides cold power by consuming electric power, decision time t k The cold power output by the electric refrigeration unit is calculated as follows:
H er (t k )＝P er (t k )η er
wherein ,Her (t k) and Per (t k ) For the decision time t k Cold power and consumed electric power output by the electric refrigerating unit; η (eta) er Is the efficiency of the electric refrigeration unit.
The scheduling process of the photovoltaic power generation device comprises the following steps:
the photovoltaic power generation device generates power by using solar energy, and the generated power can be expressed as:
wherein ,Ppv (t k) and Glight (t k ) For the decision time t k The power generation of the photovoltaic power generation device and the illumination intensity at the moment; p is p stc and Gstc The power generation power and the corresponding illumination intensity of the photovoltaic power generation device under the standard conditions.
And (3) a storage battery dispatching process:
defining a decision time t k The charge state of the storage battery is SOC es (t k ) Which represents the percentage of the remaining battery capacity to the rated capacity. The dynamic charge and discharge process of the storage battery is described as follows in consideration of the power dissipation of the storage battery and the charge and discharge in use:
SOC es (t k )＝(1-σ es )SOC es (t k -1)Δt-η es P es (t k )Δt/V es
wherein Δt is the time difference between two decision moments, σ es The energy loss rate of the storage battery is; p (P) es (t k ) For the decision time t k The storage battery charge and discharge power, positive value represents discharge, negative value represents charge, and 0 is in an idle state; v (V) es Is the capacity of the accumulator; η (eta) es The charge and discharge coefficients of the storage battery can be expressed as:
wherein , and />Is the charge and discharge efficiency of the battery.
And (3) a heat storage tank scheduling process:
defining decision time t by referring to the charge state description of the storage battery k The thermal energy storage state of the heat storage tank is SOC hs (t k ) The dynamic storage process of the heat storage tank is described as follows:
SOC hs (t k )＝(1-σ hs )SOC hs (t k -1)Δt-η hs Q hs (t k )Δt/V hs
wherein ,σhs The energy loss rate of the heat storage tank; q (Q) hs (t k ) For the decision time t k The heat storage tank stores heat release power, positive values represent heat release, negative values represent heat storage, and 0 is in an idle state; v (V) hs Is the capacity of the heat storage tank; η (eta) hs The heat storage coefficient for the heat storage tank can be expressed as:
wherein , and />The heat storage tank is used for heat storage and heat release efficiency.
Further, the method for managing the energy of the integrated energy system under the demand response is characterized in that the method for managing the energy of the integrated energy system in step 1.3 includes the following specific steps:
public service user response characteristics model:
public service users are users of office buildings, hospital buildings and the like which serve for resident production and living and have a large number of unnecessary loads, and the loads of the users can be reduced if necessary, and a load reduction model of the public service users is represented as follows:
wherein , and Pcut (t k ) For the decision time t k Original power capable of reducing load and reduced power, alpha cut (t k ) Decision time t for the user k Load-reducing ratio epsilon cut- and εcut+ The critical compensation prices a and b for which the user is willing to cut and reach the upper limit of the cut-down capability are both load cut-down coefficients.
The comprehensive energy system operator issues subsidy price to the user, the user cuts down the reducible load owned by the user according to the subsidy price, and the user makes a decision at the time t k Compensation amount C obtained after load reduction cut (t k ) Can be expressed as:
business user response characteristics model:
commercial users are some supermarkets, shopping squares and the like, which have a large amount of air conditioning loads, the modeling of the air conditioning loads comprises thermodynamic modeling of a building to which the air conditioner belongs and electric heating/cold conversion of the air conditioner, the thermodynamic model aspect of the building to which the air conditioner belongs generally adopts an equivalent thermal parameter model based on circuit simulation, and the discrete form of differential equation expression of the first-order equivalent thermal parameter model is as follows:
wherein ,Ta (t k) and To (t k ) For the decision time t k Indoor and outdoor air temperature, R a and Ca Is a roomEquivalent thermal resistance and equivalent heat capacity of internal air, P ac (t k) and Hac (t k ) For the decision time t k Electric power and refrigerating power for air conditioner, eta ac The refrigerating efficiency of the air conditioner is achieved.
Assume that the user sets the temperature to the most comfortable temperature T when not engaged in a response a,fit The decision time t can be obtained by the above equation k Air-conditioning power P for user when not participating in response ac,fit (t k ) The method comprises the following steps:
can maintain the indoor temperature within the acceptable range [ T ] of human body a,min ,T a,max ]On the premise of enabling the air conditioner to participate in scheduling, the comprehensive energy system operator and the user sign a contract to agree on a compensation price epsilon for adjusting the electric power of the air conditioner ac When necessary, the power consumption of the air conditioner is regulated, and the user decides time t k The compensation amount obtained is:
C ac (t k )＝ε ac |P ac (t k )-P ac,fit (t k )|Δt
further, the energy management method of the integrated energy system under the demand response is characterized in that the integrated energy system in step 1.4 participates in a dispatching optimization model of the power demand response, and the specific steps are as follows:
after receiving the peak regulation instruction issued by the upper power network, the comprehensive energy system operator completes a response target by making an operation plan of each device and issuing a scheduling instruction to a user on the premise of meeting the safe and stable operation of the system, and pursues the maximization of self profit R. The objective function is expressed as follows:
wherein ,Isell (t k )、I resp (t k )、C buy (t k )、C insp (t k )、C mc (t k )、C Ctax (t k ) Respectively at decision time t k The sales energy income, response income, purchase energy cost, incentive cost, maintenance fee cost and carbon tax cost of the comprehensive energy system operators; k is the total decision time number of the whole scheduling period.
The sales energy benefits are the sum of the sales electricity benefits, the sales heat benefits and the sales cooling benefits of the comprehensive energy system operators to the users, and can be expressed as:
I sell (t k )＝[p ele (t k )L ele (t k )+p hot (t k )L hot (t k )+p cold (t k )L cold (t k )]Δt
wherein ,Lele (t k )、L hot (t k) and Lcold (t k ) Respectively the decision time t k Electrical, thermal, and cold load power of the user; p is p ele (t k )、p hot (t k) and pcold (t k ) Respectively the decision time t k Electricity, heat and cold prices for the user.
The response benefit is economic compensation obtained after the comprehensive energy system responds to the power demand of the power grid, and can be expressed as follows:
I resp (t k )＝γ(t k )ε peak (t k )|P peak (t k )|Δt
wherein ,γ(tk) and εpeak (t k ) For the decision time t k Load response factor and compensation price of power demand response, gamma (t k ) And decision time t k Is related to the load response rate of (a); p (P) peak (t k ) For the decision time t k The actual response power of the comprehensive energy system is positive, which represents valley filling and negative represents peak clipping.
The energy purchase cost is the sum of the energy and natural gas energy costs purchased by the integrated energy system operator from the power grid and the natural gas grid, and can be expressed as:
C buy (t k )＝p ele (t k )P grid (t k )Δt+p gas (t k )V gas (t k )
wherein ,Pgrid (t k) and Vgas (t k ) For the system at decision time t k Purchased electric power and natural gas volume, p ele (t k) and pgas (t k ) For the decision time t k External electricity prices and natural gas prices.
The incentive cost is the total compensation cost given by the integrated energy system operator to the users who participate in the integrated demand response, and can be expressed as:
C insp (t k )＝C cut (t k )+C ac (t k )
The maintenance cost is the maintenance cost of the comprehensive energy system operator when various internal devices are operated, and can be expressed as:
wherein ,cmc,n and Pn (t k ) Maintenance costs for the unit power of the device n and decision time t k The output of the equipment N, N is the total number of the equipment.
Carbon tax costs are environmental protection fees that are levied by environmental protection departments when the system is in operation due to pollution to the environment, and can be expressed as:
wherein ,ωCtax For carbon tax coefficient, E gas and Egrid CO as unit of natural gas and electrical energy 2 Discharge quantity eta grid Is the transmission efficiency of the power grid.
Further, the comprehensive energy system energy management method under demand response is characterized in that the step 3 is to use TD3 to implement iterative updating and optimization of the policy network, the value network 1 and the value network 2, and the specific steps are as follows:
step 3.1, initializing learning and decision parameters, including: initializing the number K of decision time within one day; STEP number STEP is completed in initial learning; initializing a learning step number step=0; initializing sample pool capacity and learning sample Batch as M and Batch; initializing a discount factor gamma and a soft update coefficient tau; initializing a network parameter updating period; initializing a neural network, comprising: initializing parameters θ of policy network, value network 1 and value network 2, respectively π 、 and />Initializing parameters θ of target policy network, target value network 1, and target value network 2 π′ ←θ π 、/> and />
Step 3.2, initializing the state of the comprehensive energy system at the decision time k=0, processing the system operation data through an energy management center, and storing the generated sample into an experience pool:
step 3.2.1 bringing the state s k Normalizing to obtainWill->Inputting the current policy network to obtain corresponding action pi(s) k |θ π ) Superimposed noise v 1 Obtaining random action a k The method comprises the following steps: a, a k ＝π(s k |θ π )+v 1 Executing the currently selected action a k Through a decision periodAfter the system reaches the next state s k+1 And observe the running cost r in the process k Obtaining a sample [ s ] k ,a k ,r k ,s k+1 ]Storing the obtained product into an experience pool after normalization treatment;
step 3.2.2, returning to step 3.2.1 if k=k+1, and executing step 3.3 if k=k;
step 3.3, randomly selecting four-tuple sample data with Batch number of Batch from the experience pool
Step 3.4, obtaining the status through the target policy networkLower target action->To increase the robustness of the training process, the motion noise is superimposed on the target motion>Get random target action->Namely:
step 3.5, calculating the state through the value networkLower action->Corresponding cost function- >And
step 3.6, obtaining the in-state through the target value networkLower random target action->Corresponding objective cost function-> and />According to the Belman equation, the state is determined>Lower action->Corresponding objective cost function Q target The method comprises the following steps: />
Step 3.7 updating the value network parameters by minimizing the loss function, with respect to the parametersIs a loss function of (2)Can be expressed as: />
Step 3.8, let step=step+1, if step=n l *Cycle,N l ∈Z + Then the policy network is further modified by deterministic policy gradientsNew, it can be expressed as:
parameter θ of target policy network, target value network 1 and target value network 2 π′ 、 and />The parameters from the policy network, the value network 1 and the value network 2 are obtained by soft update and can be expressed as: θ π′ ＝(1-τ)θ π′ +τθ π 、 If STEP < STEP, return to STEP 3.2, if step=step, stop learning, network training is completed.
Compared with the prior art, the technical scheme has the following beneficial effects:
1. the invention takes the randomness of photovoltaic, temperature and electric heating and cooling load of a user into consideration while carrying out energy management on the comprehensive energy system, takes the temperature and subsidy price as factors influencing the response capability and the response willingness of the user, and based on the factors, the system can fully excavate the response potential of the user and the adjustable capability of energy conversion equipment and energy storage equipment in the system.
2. The TD3 selected by the invention is an effective deep reinforcement learning algorithm, has strong autonomous learning capability, is suitable for a high-dimensional continuous action space, avoids the problem of Q value overestimation in an actor-critic framework algorithm, and improves the training speed and stability of the algorithm.
3. The invention constructs the energy management method of the comprehensive energy system for participating in the power demand response by taking the load distribution of various users in the comprehensive energy system, the energy conversion equipment and the production plan of the energy storage equipment as decision variables together, is favorable for safely, stably and economically arranging the energy distribution in the comprehensive energy system, and further improves the potential of the comprehensive energy system for participating in the power demand response, thereby safely and efficiently completing the demand response plan of the power grid.
Drawings
FIG. 1 is a flow chart of integrated energy system energy management and control;
FIG. 2 is a block diagram of integrated energy system components and energy management;
FIG. 3 is a schematic diagram of the energy flow of the integrated energy system;
FIG. 4 is a comprehensive energy system energy management algorithm framework based on TD 3.
Detailed Description
In order to describe the technical content, constructional features, achieved objects and effects of the technical solution in detail, the following description is made in connection with the specific embodiments in conjunction with the accompanying drawings.
Example 1
Referring to fig. 2, the integrated energy system includes an energy management center (EMS), an energy conversion device including a Gas Turbine (GT), a waste Heat Boiler (HB), a Gas Boiler (GB), an Absorption Refrigerator (AR), an electric heat pump (EB), an electric refrigerator set (ER), and a photovoltaic power generation device (PV), an energy storage device, and various kinds of users, the energy conversion device being connected to a power grid through an electric bus. The energy storage equipment comprises a storage battery (ES) and a heat storage tank (HS), the multiple users comprise public service users and commercial users, the loads of the users are in three energy forms of electric load, heat load and cold load, the energy management center is an energy and information integration center, the energy and information integration center is used for externally receiving information such as an electric power supplier and a natural gas supplier, the information such as electricity price, natural gas price and a demand response plan issued by an electric company can be obtained, the purchase quantity of electric energy and natural gas energy is controlled, response offers can be issued to the multiple users in pairs to adjust the inherent energy utilization behavior of the users and the production plan of energy equipment in a control system, so that the electric power demand response of the electric company is participated and completed, and the safety and the economical efficiency of the system operation are simultaneously considered.
Referring to fig. 1, a method for managing energy of an integrated energy system under demand response includes the following steps:
step 1: the integrated energy system participates in the power demand response by taking one day as a scheduling period, and the total aim of the optimized operation is to make the total profit of the system operation for one day maximum on the basis of completing the power demand response aim, so that an operation optimization model of the integrated energy system participating in the power demand response needs to be established:
step 1.1, establishing a power demand response plan, photovoltaic output randomness, outdoor temperature randomness and a user electric heating and cooling load randomness model;
determining an amount of response to participate in the power demand response:
determining each decision time t of comprehensive energy system in one scheduling period k Response amount P 'of the power demand response of (a)' peak (t k ) Obtaining a demand response target of the comprehensive energy system at each decision moment in a scheduling period:
{P′ peak (t 1 ),P′ peak (t 2 ),...,P′ peak (t k ),...,P′ peak (t K )}
photovoltaic output randomness:
error Δp of photovoltaic output pv Mean value of compliance mu pv Variance is sigma pv Normal distribution of (a), i.e. ΔP pv ～N(μ pv ,σ pv 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the photovoltaic output is P pv,0 (t k ) The actual value P of the photovoltaic output can be obtained after the random error is superimposed pv (t k ) The method comprises the following steps:
P pv (t k )＝P pv,0 (t k )+ΔP pv
temperature randomness:
error delta T of temperature temp Mean value of compliance mu temp Variance is sigma temp Normal distribution of (a), i.e. delta T temp ～N(μ temp ,σ temp 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the temperature is T temp,0 (t k ) The actual value T of the temperature can be obtained after the random error is superimposed temp (t k ) The method comprises the following steps:
T temp (t k )＝T temp,0 (t k )+ΔT temp
electrical load randomness:
error Δl of electrical load ele Mean value of compliance mu ele Variance is sigma ele Normal distribution of (L), i.e. DeltaL ele ～N(μ ele ,σ ele 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the electric load is L ele,0 (t k ) The actual value L of the electric load can be obtained after the random error is superimposed ele (t k ) The method comprises the following steps:
L ele (t k )＝L ele,0 (t k )+ΔL ele
heat load randomness:
error Δl of thermal load hot Mean value of compliance mu hot Variance is sigma hot Normal distribution of (L), i.e. DeltaL hot ～N(μ hot ,σ hot 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the heat load is L hot,0 (t k ) The actual value L of the heat load can be obtained after the random errors are superimposed hot (t k ) The method comprises the following steps:
L hot (t k )＝L hot,0 (t k )+ΔL hot
cold load randomness:
error Δl of cooling load cold Mean value of compliance mu cold Variance is sigma cold Normal distribution of (L), i.e. DeltaL cold ～N(μ cold ,σ cold 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the cooling load is L cold,0 (t k ) The actual value L of the cold load can be obtained after the random errors are superimposed cold (t k ) The method comprises the following steps:
L cold (t k )＝L cold,0 (t k )+ΔL cold
step 1.2, establishing an energy conversion device and energy storage device scheduling model of the comprehensive energy system; referring to fig. 3, a process of energy flow in a comprehensive energy system is shown, wherein the comprehensive energy system needs to schedule its energy conversion devices and energy storage devices on the basis of maintaining a supply balance of electric load, thermal load and cold load.
Gas turbine scheduling process:
gas turbines provide electrical and thermal energy by consuming natural gas, decision time t k The power output from the gas turbine is calculated as follows:
P gt (t k )＝V gt (t k )H ng η gt
wherein ,Pgt (t k) and Vgt (t k ) For the decision time t k Electric power output by the gas turbine and consumed natural gas; h ng Is the heat value of natural gas; η (eta) gt Is the power generation efficiency of the gas turbine.
The ratio of the output thermal power to the electric power of the gas turbine is a thermoelectric ratio b, which can be expressed as:
wherein ,Qgt (t k ) For the decision time t k Heating power of the gas turbine.
The waste heat boiler scheduling process comprises the following steps:
the waste heat boiler can improve the utilization efficiency of energy, and provides heat energy for users by absorbing waste heat in the high-temperature flue gas discharged by the gas turbine, and the mathematical model is as follows:
Q hb (t k )＝Q hb,0 (t k )η hb
wherein ,Qhb (t k) and Qhb,0 (t k ) Decision time t k The heat power absorbed and output by the waste heat boiler; η (eta) hb Is the waste heat recovery efficiency.
The gas boiler scheduling process comprises the following steps:
the gas-fired boiler consumes natural gas to provide heat energy, and the gas-fired boiler makes a decision at time t k The output thermal power is calculated as follows:
Q gb (t k )＝V gb (t k )H ng η gb
wherein ,Qgb (t k) and Vgb (t k ) For the decision time t k The heat power output by the gas boiler and the consumed natural gas; η (eta) gb Is the heat generating efficiency of the gas boiler.
Scheduling process of absorption refrigerator:
the absorption refrigerator provides cold power by absorbing hot power, and at decision time t k The output cold power is calculated as follows:
H ar (t k )＝Q ar (t k )η ar
wherein ,Har (t k) and Qar (t k ) For time t k The cold power and the absorbed hot power output by the absorption refrigerator; η (eta) ar Is the refrigerating efficiency of the absorption refrigerator.
And (3) an electric heating pump scheduling process:
the electric heat pump converts low-grade heat energy into high-grade heat energy by consuming electric energy, and at decision time t k The thermal power output by the electric heat pump is calculated as follows:
Q eb (t k )＝P eb (t k )η eb
wherein ,Qeb (t k) and Peb (t k ) For the decision time t k The heat power output by the electric heat pump and the consumed electric power; η (eta) eb The heating efficiency of the electric heat pump is achieved.
Scheduling process of electric refrigerating unit:
the electric refrigerating unit provides cold power by consuming electric power, decision time t k The cold power output by the electric refrigeration unit is calculated as follows:
H er (t k )＝P er (t k )η er
wherein ,Her (t k) and Per (t k ) For the decision time t k Cold power and consumed electric power output by the electric refrigerating unit; η (eta) er Is the efficiency of the electric refrigeration unit.
The scheduling process of the photovoltaic power generation device comprises the following steps:
the photovoltaic power generation device generates power by using solar energy, and the generated power can be expressed as:
wherein ,Ppv (t k) and Glight (t k ) For the decision time t k Power generation of photovoltaic power generation device and illumination intensity at the timeA degree; p is p stc and Gstc The power generation power and the corresponding illumination intensity of the photovoltaic power generation device under the standard conditions.
And (3) a storage battery dispatching process:
Defining a decision time t k The charge state of the storage battery is SOC es (t k ) Which represents the percentage of the remaining battery capacity to the rated capacity. The dynamic charge and discharge process of the storage battery is described as follows in consideration of the power dissipation of the storage battery and the charge and discharge in use:
SOC es (t k )＝(1-σ es )SOC es (t k -1)Δt-η es P es (t k )Δt/V es
wherein Δt is the time difference between two decision moments, σ es The energy loss rate of the storage battery is; p (P) es (t k ) For the decision time t k The storage battery charge and discharge power, positive value represents discharge, negative value represents charge, and 0 is in an idle state; v (V) es Is the capacity of the accumulator; η (eta) es The charge and discharge coefficients of the storage battery can be expressed as:
wherein , and />Is the charge and discharge efficiency of the battery.
And (3) a heat storage tank scheduling process:
defining decision time t by referring to the charge state description of the storage battery k The thermal energy storage state of the heat storage tank is SOC hs (t k ) The dynamic storage process of the heat storage tank is described as follows:
SOC hs (t k )＝(1-σ hs )SOC hs (t k -1)Δt-η hs Q hs (t k )Δt/V hs
wherein ,σhs The energy loss rate of the heat storage tank; q (Q) hs (t k ) For the decision time t k The heat storage tank stores heat release power, positive values represent heat release, negative values represent heat storage, and 0 is in an idle state; v (V) hs Is the capacity of the heat storage tank; η (eta) hs The heat storage coefficient for the heat storage tank can be expressed as:
wherein , and />The heat storage tank is used for heat storage and heat release efficiency.
Step 1.3, establishing a multi-type user response characteristic model in the comprehensive energy system;
Public service user response characteristics model:
public service users are users of office buildings, hospital buildings and the like which serve for resident production and living and have a large number of unnecessary loads, and the loads of the users can be reduced if necessary, and a load reduction model of the public service users is represented as follows:
/>
wherein , and Pcut (t k ) For the decision time t k Original power capable of reducing load and reduced power, alpha cut (t k ) Decision making for a userTime t k Load-reducing ratio epsilon cut- and εcut+ For critical compensation prices that the user would like to cut and reach the upper limit of cut-off capability, a and b are both load cut-off coefficients.
The comprehensive energy system operator issues subsidy price to the user, the user cuts down the reducible load owned by the user according to the subsidy price, and the user makes a decision at the time t k Compensation amount C obtained after load reduction cut (t k ) Can be expressed as:
business user response characteristics model:
commercial users are some supermarkets, shopping squares and the like, which have a large amount of air conditioning loads, the modeling of the air conditioning loads comprises thermodynamic modeling of a building to which the air conditioner belongs and electric heating/cold conversion of the air conditioner, the thermodynamic model aspect of the building to which the air conditioner belongs generally adopts an equivalent thermal parameter model based on circuit simulation, and the discrete form of differential equation expression of the first-order equivalent thermal parameter model is as follows:
wherein ,Ta (t k) and To (t k ) For the decision time t k Indoor and outdoor air temperature, R a and Ca Is equivalent thermal resistance and equivalent heat capacity of indoor air, P ac (t k) and Hac (t k ) For the decision time t k Electric power and refrigerating power for air conditioner, eta ac The refrigerating efficiency of the air conditioner is achieved.
Assume that the user sets the temperature to the most comfortable temperature T when not engaged in a response a,fit The decision time t can be obtained by the above equation k Air-conditioning power P for user when not participating in response ac,fit (t k ) The method comprises the following steps:
can maintain the indoor temperature within the acceptable range [ T ] of human body a,min ,T a,max ]On the premise of enabling the air conditioner to participate in scheduling, the comprehensive energy system operator and the user sign a contract to agree on a compensation price epsilon for adjusting the electric power of the air conditioner ac When necessary, the power consumption of the air conditioner is regulated, and the user decides time t k The compensation amount obtained is:
C ac (t k )＝ε ac |P ac (t k )-P ac,fit (t k )|Δt
step 1.4, establishing a dispatching optimization model of the comprehensive energy system participating in power demand response;
after receiving the peak regulation instruction issued by the upper power network, the comprehensive energy system operator completes a response target by making an operation plan of each device and issuing a scheduling instruction to a user on the premise of meeting the safe and stable operation of the system, and pursues the maximization of self profit R. The objective function is expressed as follows:
wherein ,Isell (t k )、I resp (t k )、C buy (t k )、C insp (t k )、C mc (t k )、C Ctax (t k ) Respectively at decision time t k The sales energy income, response income, purchase energy cost, incentive cost, maintenance fee cost and carbon tax cost of the comprehensive energy system operators; k is the total decision time number of the whole scheduling period.
The sales energy benefits are the sum of the sales electricity benefits, the sales heat benefits and the sales cooling benefits of the comprehensive energy system operators to the users, and can be expressed as:
I sell (t k )＝[p ele (t k )L ele (t k )+p hot (t k )L hot (t k )+p cold (t k )L cold (t k )]Δt
wherein ,Lele (t k )、L hot (t k) and Lcold (t k ) Respectively the decision time t k Electrical, thermal, and cold load power of the user; p is p ele (t k )、p hot (t k) and pcold (t k ) Respectively the decision time t k Electricity, heat and cold prices for the user.
The response benefit is economic compensation obtained after the comprehensive energy system responds to the power demand of the power grid, and can be expressed as follows:
I resp (t k )＝γ(t k )ε peak (t k )|P peak (t k )|Δt
wherein ,γ(tk) and εpeak (t k ) For the decision time t k Load response factor and compensation price of power demand response, gamma (t k ) And decision time t k Is related to the load response rate of (a); p (P) peak (t k ) For the decision time t k The actual response power of the comprehensive energy system is positive, which represents valley filling and negative represents peak clipping.
The energy purchase cost is the sum of the energy and natural gas energy costs purchased by the integrated energy system operator from the power grid and the natural gas grid, and can be expressed as:
C buy (t k )＝p ele (t k )P grid (t k )Δt+p gas (t k )V gas (t k )
wherein ,Pgrid (t k) and Vgas (t k ) For the system at decision time t k Purchased electric power and natural gas volume, p ele (t k) and pgas (t k ) For the decision time t k External electricity prices and natural gas prices.
The incentive cost is the total compensation cost given by the integrated energy system operator to the users who participate in the integrated demand response, and can be expressed as:
C insp (t k )＝C cut (t k )+C ac (t k )
the maintenance cost is the maintenance cost of the comprehensive energy system operator when various internal devices are operated, and can be expressed as:
wherein ,cmc,n and Pn (t k ) Maintenance costs for the unit power of the device n and decision time t k The output of the equipment N, N is the total number of the equipment.
Carbon tax costs are environmental protection fees that are levied by environmental protection departments when the system is in operation due to pollution to the environment, and can be expressed as:
wherein ,ωCtax For carbon tax coefficient, E gas and Egrid CO as unit of natural gas and electrical energy 2 Discharge quantity eta grid Is the transmission efficiency of the power grid.
Step 2: deep neural networks required to build the dual delay depth deterministic strategy gradient algorithm (Twin Delayed Deep Deterministic Policy Gradient Algorithm, TD 3):
the TD3 is provided with three independent neural networks, namely a strategy network, a value network 1 and a value network 2, each network is provided with a respective target network, namely a target strategy network, a target value network 1 and a target value network 2, and the six neural networks are all fully connected neural networks and comprise an input layer, a hidden layer and an output layer, wherein the strategy network and the target strategy network have the same structure, and the value network 1, the value network 2, the target value network 1 and the target value network 2 have the same structure;
Step 3: referring to fig. 4, an energy management scheme of a comprehensive energy system based on a TD3 algorithm is disclosed, according to the model established in step 1, the energy management center is utilized to interact with the comprehensive energy system to obtain historical interaction information and store the historical interaction information into an experience pool, and iterative updating and optimization of a strategy network, a value network 1 and a value network 2 are realized by using TD 3;
step 3.1, initializing learning and decision parameters, including: initializing the number K of decision time within one day; STEP number STEP is completed in initial learning; initializing a learning step number step=0; initializing sample pool capacity and learning sample Batch as M and Batch; initializing a discount factor gamma and a soft update coefficient tau; initializing a network parameter updating period; initializing a neural network, comprising: initializing parameters θ of policy network, value network 1 and value network 2, respectively π 、 and />Initializing parameters θ of target policy network, target value network 1, and target value network 2 π′ ←θ π 、/> and />
Step 3.2, initializing the state of the comprehensive energy system at the decision time k=0, processing the system operation data through an energy management center, and storing the generated sample into an experience pool:
step 3.2.1 bringing the state s k Normalizing to obtain Will->Inputting the current policy network to obtain corresponding action pi(s) k |θ π ) Superimposed noise v 1 Obtaining random action a k The method comprises the following steps: a, a k ＝π(s k |θ π )+v 1 Executing the currently selected action a k After a decision period the system reaches the next state s k+1 And observe the running cost r in the process k Obtaining a sample [ s ] k ,a k ,r k ,s k+1 ]Storing the obtained product into an experience pool after normalization treatment;
step 3.2.2, returning to step 3.2.1 if k=k+1, and executing step 3.3 if k=k;
step 3.3, randomly selecting four-tuple sample data with Batch number of Batch from the experience pool
Step 3.4, obtaining the status through the target policy networkLower target action->To increase the robustness of the training process, the motion noise is superimposed on the target motion>Get random target action->Namely:
step 3.5, calculating the state through the value networkLower action->Corresponding cost function->And
step 3.6, obtaining the in-state through the target value networkLower random target action->Corresponding objective cost function-> and />According to the Belman equation, the state is determined>Lower action->Corresponding objective cost function Q target The method comprises the following steps: />
Step 3.7 updating the value network parameters by minimizing the loss function, with respect to the parametersIs a loss function of (2) Can be expressed as: />
Step 3.8, let step=step+1, if step=n l *Cycle,N l ∈Z + Then go through deterministic policyThe slight gradient updates the policy network, which can be expressed as:
parameter θ of target policy network, target value network 1 and target value network 2 π′ 、 and />The parameters from the policy network, the value network 1 and the value network 2 are obtained by soft update and can be expressed as: θ π′ ＝(1-τ)θ π′ +τθ π 、 If STEP < STEP, return to STEP 3.2, if step=step, stop learning, network training is completed.
Step 4: and 3, controlling the comprehensive energy system according to the strategy network obtained in the step 3:
by utilizing the feasibility of the established comprehensive energy system verification algorithm, at any decision time of system operation, information such as decision time, offer response quantity, operation state of the comprehensive energy system and the like is normalized and then is input into a strategy network of TD3, action information output by the strategy network is reversely normalized after forward propagation, and then actions which can be taken by the system at the current decision time can be obtained, and the situation that the system power demand response is completed after the comprehensive energy system repeatedly executes the operation in a scheduling period and the obtained selling energy benefit, response income, energy purchasing cost, excitation cost, maintenance cost and carbon tax cost are observed.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the statement "comprising … …" or "comprising … …" does not exclude the presence of additional elements in a process, method, article or terminal device comprising the element. Further, herein, "greater than," "less than," "exceeding," and the like are understood to not include the present number; "above", "below", "within" and the like are understood to include this number.
While the embodiments have been described above, other variations and modifications will occur to those skilled in the art once the basic inventive concepts are known, and it is therefore intended that the foregoing description and drawings illustrate only embodiments of the invention and not limit the scope of the invention, and it is therefore intended that the invention not be limited to the specific embodiments described, but that the invention may be practiced with their equivalent structures or with their equivalent processes or with their use directly or indirectly in other related fields.
Claims (6)
1. The comprehensive energy system energy management method under the demand response comprises a comprehensive energy system, wherein the comprehensive energy system comprises an energy management center, energy conversion equipment, energy storage equipment and multiple types of users, the energy conversion equipment comprises a gas turbine, a waste heat boiler, a gas boiler, an absorption refrigerator, an electric heat pump, an electric refrigerating unit and a photovoltaic power generation device, the energy storage equipment comprises a storage battery and a heat storage tank, the multiple types of users comprise public service users and commercial users, the loads of the users are three energy forms of electric load, thermal load and cold load, the energy management center is an integration center of energy and information, externally receives information of an electric power supplier and a natural gas supplier, can acquire information of electricity price, natural gas price, a demand response plan issued by an electric power company and the like, controls the purchase quantity of electric energy and natural gas energy, and can issue response offers to the multiple types of users to adjust the user inherent energy utilization behavior and the production plan of the energy equipment in the control system, and the energy management method is characterized by comprising the following steps:
step 1: the comprehensive energy system participates in the power demand response by taking one day as a scheduling period, the total goal of the optimized operation is to make the total profit of the system operation for one day maximum on the basis of completing the power demand response goal, and an operation optimization model of the comprehensive energy system participating in the power demand response is established:
Step 1.1, establishing a power demand response plan, photovoltaic output randomness, outdoor temperature randomness and a user electric heating and cooling load randomness model;
step 1.2, establishing an energy conversion device and energy storage device scheduling model of the comprehensive energy system;
step 1.3, establishing a multi-type user response characteristic model in the comprehensive energy system;
step 1.4, establishing a dispatching optimization model of the comprehensive energy system participating in power demand response;
step 2: the depth neural network required by the dual-delay depth deterministic strategy gradient algorithm is constructed:
the dual-delay depth deterministic strategy gradient algorithm is provided with three independent neural networks, namely a strategy network, a value network 1 and a value network 2, each network is provided with a respective target network, namely a target strategy network, a target value network 1 and a target value network 2, the six neural networks are all fully connected neural networks and comprise an input layer, a hidden layer and an output layer, the strategy network and the target strategy network have the same structure, and the value network 1, the value network 2, the target value network 1 and the target value network 2 have the same structure;
step 3: according to the model established in the step 1, the energy management center is utilized to interact with the comprehensive energy system, history interaction information is obtained and stored in an experience pool, and the TD3 is utilized to realize iterative updating and optimization of the strategy network, the value network 1 and the value network 2;
Step 4: and 3, controlling the comprehensive energy system according to the strategy network obtained in the step 3:
at any decision time of system operation, information such as decision time, offer response quantity and operation state of the comprehensive energy system is normalized and then is input into a strategy network of a dual-delay depth deterministic strategy gradient algorithm, action information output by the strategy network is inversely normalized after forward propagation, actions which can be taken by the system at the current decision time can be obtained, and the comprehensive energy system is observed to repeatedly execute the system power demand response completion condition and the obtained selling energy benefit, response benefit, energy purchasing cost, excitation cost, maintenance cost and carbon tax cost in a scheduling period.
2. The method for energy management of integrated energy systems in response to demand according to claim 1, wherein the power demand response plan, the photovoltaic output randomness, the outdoor temperature randomness, and the user electric heating and cooling load randomness model in step 1.1 are as follows:
determining an amount of response to participate in the power demand response:
determining each decision time t of comprehensive energy system in one scheduling period k Response amount P 'of the power demand response of (a)' peak (t k ) Obtaining a demand response target of the comprehensive energy system at each decision moment in a scheduling period:
{P′ peak (t 1 ),P′ peak (t 1 ),...,P′ peak (t k ),...,P′ peak (t K )}
photovoltaic output randomness:
error Δp of photovoltaic output pv Mean value of compliance mu pv Variance is sigma pv Normal distribution of (a), i.e. ΔP pv ～N(μ pv ,σ pv 2 ) The probability density function of which can representThe method comprises the following steps:
if decision time t k The predicted value of the photovoltaic output is P pv,0 (t k ) The actual value P of the photovoltaic output can be obtained after the random error is superimposed pv (t k ) The method comprises the following steps:
P pv (t k )＝P pv,0 (t k )+ΔP pv
temperature randomness:
error delta T of temperature temp Mean value of compliance mu temp Variance is sigma temp Normal distribution of (a), i.e. delta T temp ～N(μ temp ,σ temp 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the temperature is T temp,0 (t k ) The actual value T of the temperature can be obtained after the random error is superimposed temp (t k ) The method comprises the following steps:
T temp (t k )＝T temp,0 (t k )+ΔT temp
electrical load randomness:
error Δl of electrical load ele Mean value of compliance mu ele Variance is sigma ele Normal distribution of (L), i.e. DeltaL ele ～N(μ ele ,σ ele 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the electric load is L ele,0 (t k ) The electric load can be obtained after the random error is superimposedValue of L ele (t k ) The method comprises the following steps:
L ele (t k )＝L ele,0 (t k )+ΔL ele
heat load randomness:
error Δl of thermal load hot Mean value of compliance mu hot Variance is sigma hot Normal distribution of (L), i.e. DeltaL hot ～N(μ hot ,σ hot 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the heat load is L hot,0 (t k ) The actual value L of the heat load can be obtained after the random errors are superimposed hot (t k ) The method comprises the following steps:
L hot (t k )＝L hot,0 (t k )+ΔL hot
cold load randomness:
error Δl of cooling load cold Mean value of compliance mu cold Variance is sigma cold Normal distribution of (L), i.e. DeltaL cold ～N(μ cold ,σ cold 2 ) The probability density function thereof can be expressed as:
if decision time t k The predicted value of the cooling load is L cold,0 (t k ) The actual value L of the cold load can be obtained after the random errors are superimposed cold (t k ) The method comprises the following steps:
L cold (t k )＝L cold,0 (t k )+ΔL cold
3. the method for energy management of an integrated energy system in response to a demand according to claim 2, wherein the integrated energy system energy conversion device and energy storage device scheduling model of step 1.2 comprises the following specific steps:
gas turbine scheduling process:
gas turbines provide electrical and thermal energy by consuming natural gas, decision time t k The power output from the gas turbine is calculated as follows:
P gt (t k )＝V gt (t k )H ng η gt
wherein ,Pgt (t k) and Vgt (t k ) For the decision time t k Electric power output by the gas turbine and consumed natural gas; h ng Is the heat value of natural gas; η (eta) gt The power generation efficiency of the gas turbine;
the ratio of the output thermal power to the electric power of the gas turbine is a thermoelectric ratio b, which can be expressed as:
wherein ,Qgt (t k ) For the decision time t k Heating power of the gas turbine;
the waste heat boiler scheduling process comprises the following steps:
the waste heat boiler can improve the utilization efficiency of energy, and provides heat energy for users by absorbing waste heat in the high-temperature flue gas discharged by the gas turbine, and the mathematical model is as follows:
Q hb (t k )＝Q hb,0 (t k )η hb
wherein ,Qhb (t k) and Qhb,0 (t k ) Decision time t k The heat power absorbed and output by the waste heat boiler; η (eta) hb The waste heat recovery efficiency is achieved;
the gas boiler scheduling process comprises the following steps:
the gas-fired boiler consumes natural gas to provide heat energy, and the gas-fired boiler makes a decision at time t k The output thermal power is calculated as follows:
Q gb (t k )＝V gb (t k )H ng η gb
wherein ,Qgb (t k) and Vgb (t k ) For the decision time t k The heat power output by the gas boiler and the consumed natural gas; η (eta) gb The heat generating efficiency of the gas boiler;
scheduling process of absorption refrigerator:
the absorption refrigerator provides cold power by absorbing hot power, and at decision time t k The output cold power is calculated as follows:
H ar (t k )＝Q ar (t k )η ar
wherein ,Har (t k) and Qar (t k ) For time t k The cold power and the absorbed hot power output by the absorption refrigerator; η (eta) ar The refrigerating efficiency of the absorption refrigerator;
and (3) an electric heating pump scheduling process:
the electric heat pump converts low-grade heat energy into high-grade heat energy by consuming electric energy, and at decision time t k The thermal power output by the electric heat pump is calculated as follows:
Q eb (t k )＝P eb (t k )η eb
wherein ,Qeb (t k) and Peb (t k ) For the decision time t k The heat power output by the electric heat pump and the consumed electric power; η (eta) eb Heating efficiency of the electric heat pump;
scheduling process of electric refrigerating unit:
the electric refrigerating unit provides cold power by consuming electric power, decision time t k The cold power output by the electric refrigeration unit is calculated as follows:
H er (t k )＝P er (t k )η er
wherein ,Her (t k) and Per (t k ) For the decision time t k Cold power and consumed electric power output by the electric refrigerating unit; η (eta) er Efficiency of the electric refrigeration unit;
the scheduling process of the photovoltaic power generation device comprises the following steps:
the photovoltaic power generation device generates power by using solar energy, and the generated power can be expressed as:
wherein ,Ppv (t k) and Glight (t k ) For the decision time t k The power generation of the photovoltaic power generation device and the illumination intensity at the moment; p is p stc and Gstc The power generation power and the corresponding illumination intensity of the photovoltaic power generation device under the standard conditions;
and (3) a storage battery dispatching process:
defining a decision time t k The charge state of the storage battery is SOC es (t k ) Which represents the percentage of the remaining charge of the battery to the rated capacity, the dynamic charge and discharge process of the battery is described as follows, taking into account the charge dissipation and the charge and discharge in use of the battery:
SOC es (t k )＝(1-σ es )SOC es (t k -1)Δt-η es P es (t k )ΔtV es
wherein Δt is the time difference between two decision moments, σ es The energy loss rate of the storage battery is; p (P) es (t k ) For the decision time t k The storage battery charge and discharge power, positive value represents discharge, negative value represents charge, and 0 is in an idle state; v (V) es Is the capacity of the accumulator; η (eta) es The charge and discharge coefficients of the storage battery can be expressed as:
wherein , and />Charge and discharge efficiency for the battery;
And (3) a heat storage tank scheduling process:
defining decision time t by referring to the charge state description of the storage battery k The thermal energy storage state of the heat storage tank is SOC hs (t k ) The dynamic storage process of the heat storage tank is described as follows:
SOC hs (t k )＝(1-σ hs )SOC hs (t k -1)Δt-η hs Q hs (t k )Δt/V hs
wherein ,σhs The energy loss rate of the heat storage tank; q (Q) hs (t k ) For the decision time t k The heat storage tank stores heat release power, positive values represent heat release, negative values represent heat storage, and 0 is in an idle state; v (V) hs Is the capacity of the heat storage tank; η (eta) hs The heat storage coefficient for the heat storage tank can be expressed as:
wherein , and />The heat storage tank is used for heat storage and heat release efficiency.
4. The energy management method of integrated energy system under demand response according to claim 3, wherein the multiple types of user response characteristic models in the integrated energy system in step 1.3 specifically comprises the following steps:
public service user response characteristics model:
public service users are users of office buildings, hospital buildings and the like which serve for resident production and living and have a large number of unnecessary loads, and the loads of the users can be reduced if necessary, and a load reduction model of the public service users is represented as follows:
wherein , and Pcut (t k ) For the decision time t k Original power capable of reducing load and reduced power, alpha cut (t k ) Decision time t for the user k Load-reducing ratio epsilon cut- and εcut+ A and b are load shedding coefficients for critical compensation prices that the user is willing to cut and reach the upper limit of the shedding capability;
the comprehensive energy system operator issues subsidy price to the user, the user cuts down the reducible load owned by the user according to the subsidy price, and the user makes a decision at the time t k Compensation amount C obtained after load reduction cut (t k ) Can be expressed as:
business user response characteristics model:
commercial users are some supermarkets, shopping squares and the like, which have a large amount of air conditioning loads, the modeling of the air conditioning loads comprises thermodynamic modeling of a building to which the air conditioner belongs and electric heating/cold conversion of the air conditioner, the thermodynamic model aspect of the building to which the air conditioner belongs generally adopts an equivalent thermal parameter model based on circuit simulation, and the discrete form of differential equation expression of the first-order equivalent thermal parameter model is as follows:
wherein ,Ta (t k) and To (t k ) For the decision time t k Indoor and outdoor air temperature, R a and Ca Is equivalent thermal resistance and equivalent heat capacity of indoor air, P ac (t k) and Hac (t k ) For the decision time t k Electric power and refrigerating power for air conditioner, eta ac The refrigerating efficiency of the air conditioner is achieved;
assume that the user sets the temperature to the most comfortable temperature T when not engaged in a response a,fit The decision time t can be obtained by the above equation k Air-conditioning power P for user when not participating in response ac,fit (t k ) The method comprises the following steps:
can maintain the indoor temperature within the acceptable range [ T ] of human body a,min ,T a,max ]On the premise of enabling the air conditioner to participate in scheduling, the comprehensive energy system operator and the user sign a contract to agree on a compensation price epsilon for adjusting the electric power of the air conditioner ac When necessary, the power consumption of the air conditioner is regulated, and the user decides time t k The compensation amount obtained is:
C ac (t k )＝ε ac |P ac (t k )-P ac,fit (t k )|Δt。
5. the energy management method of integrated energy system under demand response according to claim 4, wherein the integrated energy system of step 1.4 participates in the dispatching optimization model of power demand response, and the specific steps are as follows:
after receiving a peak regulation instruction issued by a superior power grid, an integrated energy system operator completes a response target by making an operation plan of each device and issuing a scheduling instruction to a user on the premise of meeting the safe and stable operation of the system, and pursues the maximization of self profit R, and an objective function is expressed as follows:
wherein ,Isell (t k )、I resp (t k )、C buy (t k )、C insp (t k )、C mc (t k )、C Ctax (t k ) Respectively at decision time t k The sales energy income, response income, purchase energy cost, incentive cost, maintenance fee cost and carbon tax cost of the comprehensive energy system operators; k is the total decision time number of the whole dispatching cycle;
The sales energy benefits are the sum of the sales electricity benefits, the sales heat benefits and the sales cooling benefits of the comprehensive energy system operators to the users, and can be expressed as:
I sell (t k )＝[p ele (t k )L ele (t k )+p hot (t k )L hot (t k )+p cold (t k )L cold (t k )]Δt
wherein ,Lele (t k )、L hot (t k) and Lcold (t k ) Respectively the decision time t k Electrical, thermal, and cold load power of the user; p is p ele (t k )、p hot (t k) and pcold (t k ) Respectively the decision time t k Selling electricity, heat and cold prices to the user;
the response benefit is economic compensation obtained after the comprehensive energy system responds to the power demand of the power grid, and can be expressed as follows:
I resp (t k )＝γ(t k )ε peak (t k )|P peak (t k )|Δt
wherein ,γ(tk) and εpeak (t k ) For the decision time t k Load response factor and compensation price of power demand response, gamma (t k ) And decision time t k Load response of (2)The stress rate is related; p (P) peak (t k ) For the decision time t k The actual response power of the comprehensive energy system is positive, which represents valley filling and negative represents peak clipping;
the energy purchase cost is the sum of the energy and natural gas energy costs purchased by the integrated energy system operator from the power grid and the natural gas grid, and can be expressed as:
C buy (t k )＝p ele (t k )P grid (t k )Δt+p gas (t k )V gas (t k )
wherein ,Pgrid (t k) and Vgas (t k ) For the system at decision time t k Purchased electric power and natural gas volume, p ele (t k) and pgas (t k ) For the decision time t k External electricity prices and natural gas prices;
the incentive cost is the total compensation cost given by the integrated energy system operator to the users who participate in the integrated demand response, and can be expressed as:
C insp (t k )＝C cut (t k )+C ac (t k )
The maintenance cost is the maintenance cost of the comprehensive energy system operator when various internal devices are operated, and can be expressed as:
wherein ,cmc,n and Pn (t k ) Maintenance costs for the unit power of the device n and decision time t k The output of the equipment N, N is the total number of the equipment;
carbon tax costs are environmental protection fees that are levied by environmental protection departments when the system is in operation due to pollution to the environment, and can be expressed as:
wherein ,ωCtax For carbon tax coefficient, E gas and Egrid CO as unit of natural gas and electrical energy 2 Discharge quantity eta grid Is the transmission efficiency of the power grid.
6. The method for energy management of integrated energy system in response to demand according to claim 1, wherein the step 3 of using TD3 to implement iterative updating and optimization of the policy network, the value network 1 and the value network 2 comprises the following specific steps:
step 3.1, initializing learning and decision parameters, including: initializing the number K of decision time within one day; STEP number STEP is completed in initial learning; initializing a learning step number step=0; initializing sample pool capacity and learning sample Batch as M and Batch; initializing a discount factor gamma and a soft update coefficient tau; initializing a network parameter updating period; initializing a neural network, comprising: initializing parameters θ of policy network, value network 1 and value network 2, respectively π 、 and />Initializing parameters θ of target policy network, target value network 1, and target value network 2 π′ ←θ π 、/> and />
Step 3.2, initializing the state of the comprehensive energy system at the decision time k=0, processing the system operation data through an energy management center, and storing the generated sample into an experience pool:
step 3.2.1 bringing the state s k Normalizing to obtainWill->Inputting the current policy network to obtain corresponding action pi(s) k |θ π ) Superimposed noise v 1 Obtaining random action a k The method comprises the following steps: a, a k ＝π(s k |θ π )+v 1 Executing the currently selected action a k After a decision period the system reaches the next state s k+1 And observe the running cost r in the process k Obtaining a sample [ s ] k ,a k ,r k ,s k+1 ]Storing the obtained product into an experience pool after normalization treatment;
step 3.2.2, returning to step 3.2.1 if k=k+1, and executing step 3.3 if k=k;
step 3.3, randomly selecting four-tuple sample data with Batch number of Batch from the experience pool
Step 3.4, obtaining the status through the target policy networkLower target action->To increase the robustness of the training process, the motion noise is superimposed on the target motion>Get random target action->Namely:
step 3.5, calculating the state through the value networkLower action->Corresponding cost function- > and />
Step 3.6, obtaining the in-state through the target value networkLower random target action->Corresponding objective cost function and />According to the Belman equation, the state is determined>Lower action->Corresponding objective cost function Q target The method comprises the following steps: />
Step 3.7 updating the value network parameters by minimizing the loss function, with respect to the parametersIs a loss function of (2)Can be expressed as: />
Step 3.8, let step=step+1, if step=n l *Cycle,N l ∈Z + Updating the policy network by deterministic policy gradients can be expressed as:
parameter θ of target policy network, target value network 1 and target value network 2 π′ 、 and />The parameters from the policy network, the value network 1 and the value network 2 are obtained by soft update and can be expressed as: θ π′ ＝(1-τ)θ π′ +τθ π 、/> If STEP < STEP, return to STEP 3.2, if step=step, stop learning, network training is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310588505.3A CN116663820A (en) | 2023-05-19 | 2023-05-19 | Comprehensive energy system energy management method under demand response |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310588505.3A CN116663820A (en) | 2023-05-19 | 2023-05-19 | Comprehensive energy system energy management method under demand response |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116663820A true CN116663820A (en) | 2023-08-29 |
Family
ID=87723514
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310588505.3A Pending CN116663820A (en) | 2023-05-19 | 2023-05-19 | Comprehensive energy system energy management method under demand response |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116663820A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117394461A (en) * | 2023-12-11 | 2024-01-12 | 中国电建集团西北勘测设计研究院有限公司 | Supply and demand cooperative regulation and control system and method for comprehensive energy system |
| CN117455183A (en) * | 2023-11-09 | 2024-01-26 | 国能江苏新能源科技开发有限公司 | Comprehensive energy system optimal scheduling method based on deep reinforcement learning |
-
2023
- 2023-05-19 CN CN202310588505.3A patent/CN116663820A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117455183A (en) * | 2023-11-09 | 2024-01-26 | 国能江苏新能源科技开发有限公司 | Comprehensive energy system optimal scheduling method based on deep reinforcement learning |
| CN117394461A (en) * | 2023-12-11 | 2024-01-12 | 中国电建集团西北勘测设计研究院有限公司 | Supply and demand cooperative regulation and control system and method for comprehensive energy system |
| CN117394461B (en) * | 2023-12-11 | 2024-03-15 | 中国电建集团西北勘测设计研究院有限公司 | Supply and demand cooperative regulation and control system and method for comprehensive energy system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110458443B (en) | Smart home energy management method and system based on deep reinforcement learning | |
| Yang et al. | Indirect multi-energy transactions of energy internet with deep reinforcement learning approach | |
| CN108734350A (en) | A kind of independent method for solving with combined dispatching of the power distribution network containing micro-capacitance sensor | |
| CN116663820A (en) | Comprehensive energy system energy management method under demand response | |
| Chen et al. | A robust optimization framework for energy management of CCHP users with integrated demand response in electricity market | |
| Li et al. | Reinforcement learning of room temperature set-point of thermal storage air-conditioning system with demand response | |
| CN104699051B (en) | A kind of temperature control device demand response control method | |
| CN112036934A (en) | Quotation method for participation of load aggregators in demand response considering thermoelectric coordinated operation | |
| CN113592365B (en) | Energy optimization scheduling method and system considering carbon emission and green electricity consumption | |
| CN109884888A (en) | A kind of more building microgrid model predictions regulation method based on non-cooperative game | |
| CN107706932A (en) | A kind of energy method for optimizing scheduling based on dynamic self-adapting fuzzy logic controller | |
| Li et al. | Optimal design for component capacity of integrated energy system based on the active dispatch mode of multiple energy storages | |
| CN116432824A (en) | Comprehensive energy system optimization method and system based on multi-target particle swarm | |
| Fan et al. | Two-layer collaborative optimization for a renewable energy system combining electricity storage, hydrogen storage, and heat storage | |
| CN114662752A (en) | Comprehensive energy system operation optimization method based on price type demand response model | |
| Jintao et al. | Optimized operation of multi-energy system in the industrial park based on integrated demand response strategy | |
| Zhang et al. | Research on scheduling control strategy of large-scale air conditioners based on electric spring | |
| Li et al. | Control method of multi-energy system based on layered control architecture | |
| Yu et al. | Optimal dispatching method for integrated energy system based on robust economic model predictive control considering source–load power interval prediction | |
| Tang et al. | Multi-objective optimal dispatch for integrated energy systems based on a device value tag | |
| CN112560160A (en) | Model and data-driven heating ventilation air conditioner optimal set temperature obtaining method and equipment | |
| CN117313992A (en) | Carbon emission factor updating method considering multi-energy-flow community load | |
| Gao et al. | Multi-energy sharing optimization for a building cluster towards net-zero energy system | |
| CN111523697A (en) | Comprehensive energy service cost allocation and pricing calculation method | |
| Zhang et al. | Low carbon multi‐objective scheduling of integrated energy system based on ladder light robust optimization |
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 |