CN113191680B - Self-adaptive virtual power plant distributed architecture and economic dispatching method thereof - Google Patents
Self-adaptive virtual power plant distributed architecture and economic dispatching method thereof Download PDFInfo
- Publication number
- CN113191680B CN113191680B CN202110559796.4A CN202110559796A CN113191680B CN 113191680 B CN113191680 B CN 113191680B CN 202110559796 A CN202110559796 A CN 202110559796A CN 113191680 B CN113191680 B CN 113191680B
- Authority
- CN
- China
- Prior art keywords
- output
- deru
- iteration
- power plant
- value
- 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.)
- Active
Links
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
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/70—Smart grids as climate change mitigation technology in the energy generation sector
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- 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 relates to the technical field of electrical engineering and automation thereof, and provides a distributed economic dispatching method which has the characteristics of self-adaptability and adaptability to mass and variability of distributed resources in a virtual power plant; the decoupling property and the layered distributed architecture of the step length calculation method can adapt to the dispersion characteristic of the distributed resources, plug and play of the distributed resources can be realized, the method has reliable convergence due to the existence of the step length upper bound theoretical value, and meanwhile, the self-adaptability of the method can effectively solve the economic scheduling problem of massive distributed resources; the light design disperses the calculation and communication pressure of the system, and the operation efficiency and the adjustment speed of the virtual power plant are obviously improved; the hosting processing method for the single point failure of the coordinator is provided, and the whole method has higher fault tolerance by combining the characteristic of supporting plug and play of the distributed resources; the updating characteristics of the parameters and the layered architecture enable the parameter information of distributed resources to be protected, and privacy is achieved.
Description
Technical Field
The invention relates to the technical field of electrical engineering and automation thereof, in particular to a self-adaptive virtual power plant distributed architecture and an economic dispatching method thereof.
Background
With the expansion of the power grid scale and the increasing permeability of distributed energy resources (Distributed Energy Resource, DER), the safe and reliable regulation and economic operation of the power system are increasingly important, and the DER is reasonably and efficiently managed and participates in the economic operation of the power system, so that the research of the power system is one of important points and difficulties. The virtual power plant (Virtual Power Plant, VPP) is an ideal configuration form for aggregating DER by advanced control, metering, communication and other technologies and management mechanisms.
The existing economic dispatching method mainly has the following defects: (1) The centralized scheduling mode needs a bottom DER unit (DERU) to upload respective parameters and constraints, and a central controller uniformly solves the parameters and constraints to obtain an optimal scheme and then sends scheduling instructions to each unit. Because the DERU has the characteristic of mass dispersion, the centralized method has high requirements on communication bandwidth and central storage device performance; single point failure of the dispatch center can also cause system paralysis. And all information of the bottom layer needs to be uploaded, which is not beneficial to privacy protection of resource property owners. (2) With the development of agent technology, some scholars have proposed a distributed economic dispatch method, and deeu with certain computing and communication capabilities is regarded as an agent. After each unit performs necessary information interaction with the neighbor units, the units perform independent decision on the obtained information, so that the autonomous decision capability and individual intelligence of each unit are brought into play, for example, the current typical method, namely a distributed economic dispatch method based on multi-agent consistency, is adopted, the output marginal cost of each agent is selected as a consistency variable, and when the increment cost consistency of each unit is met, the global economy is optimized. In the prior art, a distributed economic scheduling strategy of an electric power system based on a leader-follower is proposed, but the leader is required to collect the global active power deficiency delta P in each iteration, which brings about a centralized problem, and the iteration step size has a larger influence on convergence performance. Furthermore, when the iteration exit condition Δp=0 is satisfied and the iteration is exited, the marginal cost of each agent is not necessarily strictly equal. The prior art also proposes a fully distributed economic dispatch method that does not require a leader, but requires that the sum of the outputs at the beginning of the DERU iteration be equal to the target amount. The distributed scheduling algorithm needs the communication connection of each intelligent agent to form a strong communication graph. Therefore, in practical application, the connectivity of the system needs to be checked during initialization and when the agent exits from running, so that information island is avoided, and extra workload of the system is increased, and the plug and play of distributed resources and the online fault tolerance of the system are not facilitated. In addition, since the convergence rate of the distributed scheduling algorithm is seriously dependent on the connectivity of the communication topology structure, and the iteration number for achieving convergence is obviously increased along with the expansion of the DER scale, the application of the distributed scheduling algorithm in the background of massive distributed resources is limited. (3) For the method of forming the strong communication graph on the basis of independent of all the DERU communication, the prior literature proposes a distributed algorithm based on an Alternate Direction Multiplier Method (ADMM), information exchange among agents is not needed, information islands are effectively avoided, however, the influence of punishment coefficients in the algorithm on convergence is large, the influence is difficult to analyze in a definite manner, the punishment coefficients are generally obtained by trial and error, and the application of the punishment coefficients is limited. In addition, all the above methods rely on the fact that the cost function of the DERU is a quadratic function of the output power, limiting the versatility of the method.
Disclosure of Invention
The invention aims to establish a distributed economic dispatching method and a distributed economic dispatching architecture of a virtual power plant, which are applicable to mass distributed resources. The method has the characteristics of stable convergence condition and self-adaptive adjustment, can realize plug and play of the DERU, can solve the problem of single-point fault of a coordinator, improves the calculation and communication efficiency, protects privacy, and has stronger fault tolerance.
In order to achieve the above purpose, the invention provides a technical scheme that the self-adaptive virtual power plant distributed architecture comprises an Agent layer, a gateway layer, a database, a message queue server, a coordination layer and a power grid;
the Agent layer comprises a plurality of groups of DERUs, the power supply equipment or the adjustable load at the bottommost layer in the virtual power plant of the DERU unit has sensing and calculating capabilities, can respond to an adjusting signal according to an increment cost function of the Agent layer, executes a final output instruction, and can continuously adjust power output within an output limit value range; the DERU is connected to a power bus of the virtual power plant through an energy interface, and is in communication connection with a local energy information gateway, and one gateway can be connected with dozens of DERUs;
the gateway layer is a middle layer and has an edge calculation function, gathers and uploads information of the bottom layer DERU in each iteration, and transmits an adjusting signal issued by the coordination layer to the DERU;
an iteration management service exists in the coordination layer to advance iteration, and in a time window of each iteration, all messages are asynchronous to be published or subscribed to relieve high concurrent communication pressure;
the message queue server may write important information in the queue to a database for reading from time to time as needed.
A decentralized economic dispatch method for an adaptive virtual power plant, comprising the steps of:
s1, the power grid issues an output instruction to the virtual power plant, namely an output target quantity X D ；
S2, the coordination layer issues an initialization adjusting signal, namely an increment cost value lambda, to the gateway, and the gateway transmits the adjusting signal to the on-line DERU;
s3, the DERU calculates the self-output forceAnd sensitivity 1/lambda' i (x i (k) Comparing the output limit value with the output limit value, and reporting the output size of the DERU and the sensitivity of the DERU to the adjusting signal if the output limit value is not exceeded; if the output limit value is exceeded, the DERU output limit value is set at the output limit value, andits sensitivity is 1/lambda' i (x i (k) Reporting the output size and the sensitivity of the DERU to the gateway after zeroing;
s4, the gateway gathers the total output value and the total sensitivity value of the connected DERU and issues the total output value and the total sensitivity value to the coordination layer through the message queue;
s5, the coordination layer calculates the total output value and the step upper limit of all the DERUs, and calculates the output error delta X (k), namely the output target quantity X D A difference from the total force value;
s6, if the output error is small enough, the iteration is ended, the coordination layer issues a unified iteration ending signal, each DERU outputs according to the final iteration calculation result, and the virtual power plant completes economic dispatch; if the output error is larger, the magnitude of the adjusting signal is changed, and the step S2 is returned to for the next iteration until the error is small enough.
In step S1, when the virtual coordinator in the coordination layer fails, the computing device is reassigned to be a new virtual coordinator, and the new virtual coordinator subscribes to all the gateway-counted output values x l Step size h l Wherein, l is the gateway number;
in step S4, when the virtual coordinator in the coordination layer fails, the computing device is reassigned to be a new virtual coordinator, and the new virtual coordinator subscribes to all the gateway-counted output values x l Step size h l The new virtual coordinator downloads the output target amount X from the database or message queue D And the current iteration's adjustment signal λ (k) and the total output value and total sensitivity value of all connected DERUs issued by each gateway;
in step S6, the rule for changing the adjustment signal size is: if the total output value is lower than the output target quantity X D Increasing the adjusting signal in the next iteration, if the total output value is higher than the output target value X D The adjusting signal at the next iteration is reduced until the total output value and the output target quantity X D The difference is small enough and the iteration is ended;
in step S6, the update rule of the adjustment signal is:
wherein k is the iteration number; h is a positive scalar, which is the iteration step; x is X D Is the output target of VPP, the size isx i For the power generation of the ith DERU, x is the load-reducible power i Reducing the load thereof; x is x i The force constraint formula of (2) is +.>i=1, 2, …, N, where +.>And->The upper limit and the lower limit of the output power of the ith DERU are respectively set;
order theFor the increment cost of the ith DERU, if the force constraint formula is not considered, the increment rate is equal to the value lambda by the micro increment rate criterion 1 ＝λ 2 ＝…＝λ N ＝λ * And satisfy->When the virtual power plant has the lowest output cost, the economic dispatch is achieved; if the output constraint formula is considered, the DERU output exceeding the output limit value is required to be fixed at the corresponding limit value in the solving process, and when the increment cost of the DERU which does not reach the limit value is consistent, the virtual power plant realizes economic dispatch;
defining the error of the total output of the virtual power plant relative to the target amount in each iteration as follows:
update rule of DERU output:
that is, when the adjustment signal exceeds the dera's output range, the output will be fixed at a limit value, at which time the dera reaching the limit value will not continue to respond to changes in the adjustment signal; wherein, the liquid crystal display device comprises a liquid crystal display device,lambda is lambda i (-), an inverse function of (-).
The iteration step h has three convergence conditions, which are respectively as follows:
1) When the cost function of all the DERUs is a quadratic function
When the cost function of the DERU is a conventional quadratic function:
wherein x is i And f i The power generation and cost of the ith DERU are respectively, and for load reduction, x i And f i The costs are reduced and regulated, respectively, for their loads, while the dehu generation or regulation costs in the power system satisfy the marginal cost increment law, thus the cost function f i Is a monotonically increasing function, a i 、b i 、c i The coefficients of the quadratic term, the primary term and the constant term are positive numbers respectively, and if h is unchanged in each iteration and the output limit of the DERU is not considered, the sufficient and necessary conditions for effective convergence in the process are as follows:
this means that the adjustment step h cannot be arbitrarily valued, should be a positive number, and there is an upper bound only related to the quadratic term coefficients of the cost function of all DERU, independent of the first-order term coefficients and the constant term, so the step upper bound can be defined:
the above convergence condition becomes:
h＝rh max 0＜r＜1 (7)
the iterative convergence condition is the iterative convergence condition when the cost functions of all the DERUs are quadratic functions;
2) When the delta function of the DERU is the common delta function
Derivative of the cost function of the DERU, i.e. the incremental cost function lambda i (x i ) To increase the function, the step size upper limit is:
wherein lambda' i (x i ) Lambda is lambda i (x i ) For x i Is the first derivative of (f) i For x i Is a second derivative of (2);
3) When the DERU reaches the output limit or communication failure
When the DERU reaches the force limit, the DERU will no longer have the ability to respond to the change in the adjustment signal, thus calculating h max When the step size is needed to be ignored, the step size centralized updating rule in the iterative process is as follows: considering the output limit of the DERU, in the case that the incremental cost function of the DERU is a general incremental function, in the kth iteration, the algorithm of the centralized updating rule of the step h is:
a1：ξ←0
a2：for i＝1:N
a3：
a4：thenξ←ξ+1/λ′ i (x i (k))
a5：h max (k)←2/ξ
a6：h(k)＝rh max (k)
special case treatment: when the output value corresponding to the adjusting signal is larger than the upper limit of a part of the DERU and is lower than the lower limit of the rest of the DERU, the iteration encounters a dead zone, and the step length is updated under the condition that the denominator is 0, namely, xi=0 of the algorithm a5 line, and when the condition happens, the limit value judgment of the algorithm a4 line can only take one side, namely, the error deltax (k) is positive, only the upper limit is judged, and only the lower limit is judged when the error deltax (k) is negative; that is, the denominator ζ is determined by the DERU whose output is adjustable in the error direction.
Compared with the prior art, the invention has the beneficial effects that: the distributed economic dispatching method has the characteristics of self-adaptability and adaptability to mass and variability of distributed resources in the virtual power plant; the decoupling performance and the layered distributed architecture of the step size calculation method can adapt to the dispersion characteristic of the distributed resources, and the plug and play of the distributed resources can be realized.
1) The existence of the upper bound theoretical value of the step length ensures that the method has reliable convergence, and meanwhile, the self-adaptability of the method can effectively solve the economic scheduling problem of mass distributed resources;
2) The light design disperses the calculation and communication pressure of the system, and the operation efficiency and the adjustment speed of the virtual power plant are obviously improved;
3) The hosting processing method for the single point failure of the coordinator is provided, and the whole method has higher fault tolerance by combining the characteristic of supporting plug and play of the distributed resources;
4) The updating characteristics of the parameters and the layered architecture enable the parameter information of distributed resources to be protected, and privacy is achieved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a virtual power plant decentralized architecture based on a message queue transmission mechanism;
FIG. 2 is a flow chart of a distributed economic dispatch method for a virtual power plant.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
1. Virtual power plant economic dispatch model
The economic dispatch of the virtual power plant means that the output of each generator set is reasonably arranged under the condition of meeting the total output target of the virtual power plant and the output range of each distributed resource, so that the total power generation cost is minimum.
Wherein x is i And f i The power generation and cost of the ith DERU are respectively, and for load reduction, x i And f i Respectively reducing and adjusting costs for its load. Whereas the dehu generation or regulation costs in an electrical power system satisfy the marginal cost increasing rule, thus f i (cost function)Number) is a monotonically increasing function. N is the total number of DERUs, F is the total power generation cost of the virtual power plant, and X D As the output target amount of the VPP,and->The upper and lower limits of the output power of the ith DERU are respectively set. Let->Incremental Cost (IC) for the ith dehu. If the force constraint (1 c) is not considered, the force constraint is calculated by the equal-minute increment rule, when lambda 1 ＝λ 2 ＝…＝λ N When =λ, and (1 b) is satisfied, the output cost of VPP is the lowest, and economic dispatch is achieved. For the constraint optimization problem when considering equation (1 c), the dere output exceeding the output limit value needs to be fixed at the corresponding limit value in the solving process, and when the incremental costs of the dere which do not reach the limit value are consistent, the VPP (virtual power plant) realizes economic dispatch.
2. Virtual power plant decentralized economic dispatch method description
For a decentralized scheduling algorithm, there is no direct communication between units, so an upper layer body is required to coordinate the iteration directions of the units, and the method defines the coordination role as a virtual coordinator (virtual coordinator, VC). And because the distributed architecture design distributes the calculation pressure, compared with the traditional central controller, the performance requirement of the coordinator can be greatly reduced, and the coordinator can be born by a local energy information gateway. Thus, the method weakens the functional requirements for the VC through proper architecture and flow design, so that any (or most) intermediate layer body with edge intelligence can take the role of coordinator so that all units overall optimally fulfill one overall task. In addition, because the system is light in weight and has low performance requirement on the VC, when the VC fails, coordination function hosting can be conveniently carried out, and certain fault tolerance can be realized.
Equal nuances to meet economic dispatchIn each iteration, the VC issues a unified incremental cost value λ as an adjustment signal, from which each DERU calculates its own output and reports. To satisfy the power balance constraint of equation (2), based on the feedback thought, VC will sum the total dera output and the target value X D Comparing, if it is lower than X D The adjustment signal at the next iteration is increased if it is higher than X D The adjustment signal is reduced until the total force of the DERU in VPP and the target value X D The error between the two is small enough, the iteration is ended, the VC issues a unified iteration ending signal, each DERU outputs force according to the final iteration calculation result, and the VPP completes economic dispatching. To ensure fault tolerance, the system will adaptively adjust the step size from bottom to top in each iteration according to the online condition and the running state of the DERU to ensure the convergence speed.
3. Iterative update rules for parameters
The update rule of the adjustment signal is:
wherein k is the iteration number; h is a positive scalar, which is an iteration step, has an important effect on convergence. The error of the total VPP output relative to the target amount in each iteration is defined as:
the update rule of the DERU output is:
that is, when the adjustment signal exceeds the DERU's force range, the force will be fixed at a limit value, at which time the DERU reaching the limit value will not continue to respond to changes in the adjustment signal. Wherein, the liquid crystal display device comprises a liquid crystal display device,lambda is lambda i (-), an inverse function of (-).
4. Convergence condition
(1) The cost function of all the DERUs is a quadratic function
When the cost function of the DERU is a conventional quadratic function:
time (a) i 、b i 、c i The quadratic term, the primary term coefficient and the constant term are positive numbers respectively), if h does not change in each iteration and the output limit of the DERU is not considered, the sufficient and necessary conditions for effective convergence in the above process are:
this means that the adjustment step h cannot be arbitrarily chosen, should be positive, and there is an upper bound only on the quadratic term coefficients of the cost function of all DERU, independent of the first term coefficients and the constant term. Thus, a step upper bound can be defined:
the above convergence condition becomes:
h＝rh max 0＜r＜1
the formula is the iterative convergence condition when the cost function of all the DERUs is a quadratic function.
(2) The delta function of the DERU is a common delta function
More generally, the derivative of the cost function of the DERU, i.e. the incremental cost function λ i (x i ) To increase the function, the step size upper limit is:
wherein lambda' i (x i ) Lambda is lambda i (x i ) For x i Is the first derivative of (f) i For x i Is a second derivative of (c).
(3) DERU reaches the output limit or communication failure
When the DERU reaches the force limit, the DERU will no longer have the ability to respond to the change in the adjustment signal, thus calculating h max When the step size is needed to be ignored, the step size centralized updating rule in the iterative process is as follows:
when the DERU fails in communication, the parameter value of the DERU will not be uploaded in iteration, and the iteration end signal issued by the VC will not be received finally, and the output is still performed according to the original working point without affecting the iteration process of other DERU, so that the step update rule of algorithm 1 can still ensure convergence. In a physical sense, 1/lambda' i (x i (k) In the kth iteration, the output of the ith DERU is the sensitivity of the adjustment signal. The greater the sensitivity, the greater the response of the DERU to the adjustment signal, i.e., the greater the ratio of the DERU output change micro-increment to the adjustment signal change micro-increment, within the output limit. In combination with algorithm 1, it can be found that the step size has an adaptive nature: the larger the number of the DERUs with response capability and the larger the output sensitivity of each DERU, the smaller the step size and the smaller the possibility of overshoot; conversely, the smaller the number of derers with response capability, the smaller the output sensitivity of each derer, the larger the step size and the less the possibility of undershooting.
Special case treatment: when the corresponding output value of the adjustment signal is greater than the upper limit of a part of the DERU and lower than the lower limit of the rest of the DERU, the iteration encounters a dead zone, and the step update encounters a case where the denominator is 0 (ζ=0 on line 5 of algorithm 1). When this occurs, the limit value determination on line 4 of algorithm 1 may take only one side, i.e., the error Δx (k) is positive, only the upper limit is determined, and when the error Δx (k) is negative, only the lower limit is determined. That is, the denominator ζ is determined by the DERU whose output is adjustable in the error direction.
Algorithm 1 describes the update rule of the step size in a centralized case, and experiments show that r=0.5 will have the fastest convergence speed. Further, the distributed virtual power plant internal architecture and calculation decoupling can be designed into a distributed step updating mode. In each iteration, the self-force x is calculated by each DERU i (k) And 1/lambda' i (x i (k) Therefore, autonomous decision making and individual intelligence of each unit are exerted, and the calculation pressure is dispersed. In addition, through the layered distributed architecture of the DERU-gateway-VC, the gateway is used as an intermediate layer individual to collect the parameter information of tens of DERUs connected with the gateway, and the added intermediate data is uploaded, so that the calculation pressure of the VC can be further weakened, the parameter information of a single DERU is shielded for the VC, and the privacy is protected. In terms of fault tolerance, when the DERU has communication faults, the reliability and the convergence speed of convergence are guaranteed by the step length calculation in the algorithm 1; when the communication failure occurs in the VC, the coordination function of the VC can be hosted on other computing devices in a hosting mode, necessary data in the iteration process are reserved, and then all gateways are in communication connection with the new VC to continue iteration. Thus, adaptation of the step size in combination with the hosting mechanism of the VC may enable high fault tolerance and high availability of the overall system.
5. Virtual power plant decentralized architecture
As can be seen from the description and convergence conditions of the decentralized economic dispatch method, the upper VC in the virtual power plant is required to count the total amount of the DERU output and necessary parameter information in each iteration, update the step length and calculate the unified regulating signal, and then issue the regulating signal. All the DERUs feed back their own output according to the adjustment signal, so as to do so until the error of the total output force of the DERUs and the output command target is within the allowable range. A virtual power plant decentralized architecture based on a message queue transmission mechanism is designed to decouple the different parts and enhance the scalability and fault tolerance of the system, as shown in fig. 1. In the virtual power plant, the DERU is the power supply equipment or adjustable load of the bottommost layer (Agent layer), has sensing and calculating capabilities, can respond to an adjusting signal according to an increment cost function of the DERU, executes a final output command, and can continuously adjust power output within an output limit value range. The DERUs are connected to the power bus of the virtual power plant through the energy interface, communication connection is established between the DERUs and a local energy information gateway, and one gateway can be connected with dozens of DERUs. The middle gateway layer has an edge calculation function, gathers and uploads information of the bottom layer DERU in each iteration, and communicates the VC published adjustment signal to the DERU. And the VC of the coordination layer is logically connected with all the gateways in a communication way, and step length and adjustment signal updating are carried out so as to coordinate the total output of the virtual power plant to finally meet the output command of the power grid. There is an iteration management service in the coordination layer to advance iterations, in each iteration's time window, all messages are published or subscribed asynchronously to alleviate the high concurrency of communication pressure. The message queue server may write important information in the queue to the database for reading when needed.
In the message queue transmission mechanism, all devices of the gateway layer and the coordination layer only need to establish connection with a message queue server, and each device subscribes to the required messages through a listener. The gateway of the gateway layer needs to subscribe the excitation signal information published by the VC, and the VC needs to subscribe the parameter summarization information of all the gateways. All the DERUs independently calculate their own forces in response to the adjustment signal and the communication and calculation pressure of the VC is relieved through the aggregation of the gateway layers. The VC uses algorithm 1 and equation (1) to update the step size and the excitation signal, and the hierarchical and decentralized architecture and communication decoupling design makes the communication and computation of the VC very light, so the VC can also be acted on by the gateway device. When single point failure occurs, the function of the VC can be managed to other gateways, and all the gateways can recover iteration only by communicating with the new VC. Due to the lightweight design, the VC may even be directly acted upon by a certain gateway of the gateway layer. When a VC fails, a legacy message may be issued to trigger a coordination function hosting mechanism. After the coordination function is hosted on other gateways or computing servers, the new VC only downloads iteration-related parameter information from the message queue or database and then continues to organize iterations. Therefore, the iteration can be recovered only by reserving the total output instruction and certain information in the iteration in the database or the message queue, and the storage pressure is low.
6. Method flow design
The basic flow of the decentralized economic dispatch method is shown in figure 2. Wherein L is the gateway number, and L is the number of gateways in the gateway layer. The number of DERUs connected with the first gateway is n l The method comprises the following steps: the output of the ith DERU connected to the ith gateway. X is X l And h l The sum of the total output of all the DERUs connected to the first gateway and the inverse of the derivative of the incremental cost function, respectively. VC needs to subscribe to X published by all gateways l And h l Information. The method decouples the step length updating calculation process, solves the VC single-point fault through a VC hosting mechanism, and improves the fault tolerance and usability of the method. When VC fails, new VC needs to download X in this iteration l 、h l Lambda and X D The VC function can then be performed by the same flow to recover the iteration, so the single point failure problem of the coordinator is solved. When the DERU or the gateway fails or is disconnected, the step length calculation can be ignored; when the DERU is added in the iteration process, the step length is updated from bottom to top to meet the convergence condition, so that the convergence performance is not affected, and the plug-and-play of the distributed resources can be realized.
The distributed architecture and the method flow of the virtual power plant have the characteristic of high decoupling, and the gateway gathers the bottom layer information and then uploads the intermediate result to the coordination layer after adding operation, so that in each iteration, a coordinator does not know specific parameters and state information of a single DERU device, and the distributed architecture and the method flow have the characteristic of privacy protection. Further, the architecture can be improved, for example, multiple layers of middle layers are designed, the main bodies of the same layer are decoupled, and the upper layer gathers information of the lower layer and uploads the information, so that communication and calculation of each device are lighter.
In the description of the present specification, the descriptions of the terms "one embodiment," "example," "specific example," and the like, mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiments or examples. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The preferred embodiments of the invention disclosed above are intended only to assist in the explanation of the invention. The preferred embodiments are not exhaustive or to limit the invention to the precise form disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best understand and utilize the invention. The invention is limited only by the claims and the full scope and equivalents thereof.
Claims (7)
1. The self-adaptive virtual power plant distributed architecture is characterized by comprising an Agent layer, a gateway layer, a database, a message queue server, a coordination layer and a power grid;
the Agent layer comprises a plurality of groups of DERUs, the DERUs are the power supply equipment or adjustable load at the bottommost layer in the virtual power plant, have sensing and calculating capabilities, can respond to adjusting signals according to self incremental cost functions, execute final output instructions, and can continuously adjust power output within an output limit value range; the DERU is connected to a power bus of the virtual power plant through an energy interface, and is in communication connection with a local energy information gateway, and one gateway can be connected with dozens of DERUs;
the gateway layer is a middle layer and has an edge calculation function, gathers and uploads information of the bottom layer DERU in each iteration, and transmits an adjusting signal issued by the coordination layer to the DERU;
an iteration management service exists in the coordination layer to advance iteration, and in a time window of each iteration, all messages are asynchronous to be published or subscribed to relieve high concurrent communication pressure;
the message queue server may write important information in the queue to a database for reading from time to time as needed.
2. A decentralized economic dispatch method for an adaptive virtual power plant according to claim 1, comprising the steps of:
s1, the power grid issues an output instruction to the virtual power plant, namely an output target quantity X D ；
S2, the coordination layer issues an initialization adjusting signal, namely an increment cost value lambda, to the gateway, and the gateway transmits the adjusting signal to the on-line DERU;
s3, the DERU calculates the self-output forceAnd sensitivity 1/lambda' i (x i (k) Comparing the output limit value with the output limit value, and reporting the output size of the DERU and the sensitivity of the DERU to the adjusting signal if the output limit value is not exceeded; if the output limit is exceeded, the DERU output limit is set at the output limit and the sensitivity is 1/lambda' i (x i (k) Reporting the output size and the sensitivity of the DERU to the gateway after zeroing;
s4, the gateway gathers the total output value and the total sensitivity value of the connected DERU and issues the total output value and the total sensitivity value to the coordination layer through the message queue;
s5, the coordination layer calculates the total output value and the step upper limit of all the DERUs, and calculates the output error delta X (k), namely the output target quantity X D A difference from the total force value;
s6, if the output error is small enough, the iteration is ended, the coordination layer issues a unified iteration ending signal, each DERU outputs according to the final iteration calculation result, and the virtual power plant completes economic dispatch; if the output error is larger, the magnitude of the adjusting signal is changed, and the step S2 is returned to for the next iteration until the error is small enough.
3. An adaptive virtual power plant decentralized economic dispatch method according to claim 2, wherein: in step S1, when the virtual coordinator in the coordination layer fails, the computing device is reassigned to be a new virtual coordinator, and the new virtual coordinator subscribes to all the gateway-counted output values x l Step size h l Where l is the gateway number.
4. A method of decentralized economic dispatch for an adaptive virtual power plant according to claim 3, wherein: in step S4, when the virtual coordinator in the coordination layer fails, the computing device is reassigned to be a new virtual coordinator, and the new virtual coordinator subscribes to all the gateway-counted output values x l Step size h l The new virtual coordinator downloads the output target amount X from the database or message queue D And the adjustment signal λ (k) of the current iteration and the total output value and the total sensitivity value of all connected DERUs issued by the respective gateway.
5. The method according to claim 4, wherein in step S6, the rule for changing the size of the adjustment signal is: if the total output value is lower than the output target quantity X D Increasing the adjusting signal in the next iteration, if the total output value is higher than the output target value X D The adjusting signal at the next iteration is reduced until the total output value and the output target quantity X D Is small enough and the iteration ends.
6. The adaptive virtual power plant decentralized economic dispatch method of claim 5, wherein:
the update rule of the adjustment signal is:
wherein k is the iteration number; h is a positive scalar, which is the iteration step; x is X D Is the output target of VPP, the size isx i For the power generation of the ith DERU, x is the load-reducible power i Reducing the load thereof; x is x i The force constraint formula of (2) isi=1, 2, …, N, where +.>And->The upper limit and the lower limit of the output power of the ith DERU are respectively set;
order theFor the increment cost of the ith DERU, if the force constraint formula is not considered, the increment rate is equal to the value lambda by the micro increment rate criterion 1 ＝λ 2 ＝…＝λ N ＝λ * And satisfy->When the virtual power plant has the lowest output cost, the economic dispatch is achieved; if the output constraint formula is considered, the DERU output exceeding the output limit value is required to be fixed at the corresponding limit value in the solving process, and when the increment cost of the DERU which does not reach the limit value is consistent, the virtual power plant realizes economic dispatch;
defining the error of the total output of the virtual power plant relative to the target amount in each iteration as follows:
update rule of DERU output:
that is, when the adjustment signal exceeds the dera's output range, the output will be fixed at a limit value, at which time the dera reaching the limit value will not continue to respond to changes in the adjustment signal; wherein lambda is i -1 (. Cndot.) is lambda i (-), an inverse function of (-).
7. The adaptive virtual power plant decentralized economic dispatch method of claim 6, wherein: the iteration step h has three convergence conditions, which are respectively as follows:
1) When the cost function of all the DERUs is a quadratic function
When the cost function of the DERU is a conventional quadratic function:
wherein x is i And f i The power generation and cost of the ith DERU are respectively, and for load reduction, x i And f i The costs are reduced and regulated, respectively, for their loads, while the dehu generation or regulation costs in the power system satisfy the marginal cost increment law, thus the cost function f i Is a monotonically increasing function, a i 、b i 、c i The coefficients of the quadratic term, the primary term and the constant term are positive numbers respectively, and if h is unchanged in each iteration and the output limit of the DERU is not considered, the sufficient and necessary conditions for effective convergence in the process are as follows:
this means that the adjustment step h cannot be arbitrarily valued, should be a positive number, and there is an upper bound only related to the quadratic term coefficients of the cost function of all DERU, independent of the first-order term coefficients and the constant term, so the step upper bound can be defined:
the above convergence condition becomes:
h＝rh max 0＜r＜1 (7)
the iterative convergence condition is the iterative convergence condition when the cost functions of all the DERUs are quadratic functions;
2) When the delta function of the DERU is the common delta function
Derivative of the cost function of the DERU, i.e. the incremental cost function lambda i (x i ) To increase the function, the step size upper limit is:
wherein lambda' i (x i ) Lambda is lambda i (x i ) For x i Is the first derivative of (f) i For x i Is a second derivative of (2);
3) When the DERU reaches the output limit or communication failure
When the DERU reaches the force limit, the DERU will no longer have the ability to respond to the change in the adjustment signal, thus calculating h max When the step size is needed to be ignored, the step size centralized updating rule in the iterative process is as follows: considering the output limit of the DERU, in the case that the incremental cost function of the DERU is a general incremental function, in the kth iteration, the algorithm of the centralized updating rule of the step h is:
a1：ξ←0
a2：for i＝1:N
a5：h max (k)←2/ξ
a6：h(k)＝rh max (k)
special case treatment: when the output value corresponding to the adjusting signal is larger than the upper limit of a part of the DERU and is lower than the lower limit of the rest of the DERU, the iteration encounters a dead zone, and the step length is updated under the condition that the denominator is 0, namely, xi=0 of the algorithm a5 line, and when the condition happens, the limit value judgment of the algorithm a4 line can only take one side, namely, the error deltax (k) is positive, only the upper limit is judged, and only the lower limit is judged when the error deltax (k) is negative; that is, the denominator ζ is determined by the DERU whose output is adjustable in the error direction.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110559796.4A CN113191680B (en) | 2021-05-21 | 2021-05-21 | Self-adaptive virtual power plant distributed architecture and economic dispatching method thereof |
| JP2021179297A JP7260618B2 (en) | 2021-05-21 | 2021-11-02 | Distributed Architecture of Self-Adapting Virtual Power Plant and its Economic Allocation Method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110559796.4A CN113191680B (en) | 2021-05-21 | 2021-05-21 | Self-adaptive virtual power plant distributed architecture and economic dispatching method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113191680A CN113191680A (en) | 2021-07-30 |
| CN113191680B true CN113191680B (en) | 2023-08-15 |
Family
ID=76984835
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110559796.4A Active CN113191680B (en) | 2021-05-21 | 2021-05-21 | Self-adaptive virtual power plant distributed architecture and economic dispatching method thereof |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JP7260618B2 (en) |
| CN (1) | CN113191680B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117239728B (en) * | 2023-09-19 | 2024-03-22 | 盐城工学院 | Distributed economic dispatching method for micro-grid |
| CN117477660A (en) * | 2023-10-16 | 2024-01-30 | 广州高新区能源技术研究院有限公司 | Soft light storage and filling system joint regulation and control method and system based on VPP (virtual private point) demand response |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103824134A (en) * | 2014-03-06 | 2014-05-28 | 河海大学 | Two-stage optimized dispatching method for virtual power plant |
| JP2018143046A (en) * | 2017-02-28 | 2018-09-13 | 株式会社ダイヘン | Virtual Power Plant |
| CN110416998A (en) * | 2019-07-01 | 2019-11-05 | 华北电力大学 | A kind of complicated distribution scheduling Control management system in area based on virtual power plant |
| CN111242443A (en) * | 2020-01-06 | 2020-06-05 | 国网黑龙江省电力有限公司 | Deep reinforcement learning-based economic dispatching method for virtual power plant in energy internet |
| CN112418636A (en) * | 2020-11-17 | 2021-02-26 | 海南省电力学校(海南省电力技工学校) | Self-organizing polymerization operation scheduling method for virtual power plant |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9960637B2 (en) * | 2015-07-04 | 2018-05-01 | Sunverge Energy, Inc. | Renewable energy integrated storage and generation systems, apparatus, and methods with cloud distributed energy management services |
| JP6963397B2 (en) * | 2017-03-03 | 2021-11-10 | 株式会社日立製作所 | Aggregation control system, aggregation control method and control device |
| US10880362B2 (en) * | 2018-12-03 | 2020-12-29 | Intel Corporation | Virtual electrical networks |
| JP7280556B2 (en) * | 2019-09-26 | 2023-05-24 | 東京電力ホールディングス株式会社 | Negawatt trading support device and negawatt trading method |
| CN111969596A (en) | 2020-08-02 | 2020-11-20 | 上海千贯节能科技有限公司 | Load self-adaptive correction response method of electrical-grade load frequency control system |
-
2021
- 2021-05-21 CN CN202110559796.4A patent/CN113191680B/en active Active
- 2021-11-02 JP JP2021179297A patent/JP7260618B2/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103824134A (en) * | 2014-03-06 | 2014-05-28 | 河海大学 | Two-stage optimized dispatching method for virtual power plant |
| JP2018143046A (en) * | 2017-02-28 | 2018-09-13 | 株式会社ダイヘン | Virtual Power Plant |
| CN110416998A (en) * | 2019-07-01 | 2019-11-05 | 华北电力大学 | A kind of complicated distribution scheduling Control management system in area based on virtual power plant |
| CN111242443A (en) * | 2020-01-06 | 2020-06-05 | 国网黑龙江省电力有限公司 | Deep reinforcement learning-based economic dispatching method for virtual power plant in energy internet |
| CN112418636A (en) * | 2020-11-17 | 2021-02-26 | 海南省电力学校(海南省电力技工学校) | Self-organizing polymerization operation scheduling method for virtual power plant |
Non-Patent Citations (1)
| Title |
|---|
| 以高压电极式电锅炉为核心的高性能分布式电热能源系统;康重庆;陈辉;赵宇明;韩兵;高峰;杨经纬;Ton Verra Huibert;John N.Jiang;;南方电网技术(第10期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113191680A (en) | 2021-07-30 |
| JP2022179294A (en) | 2022-12-02 |
| JP7260618B2 (en) | 2023-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113191680B (en) | Self-adaptive virtual power plant distributed architecture and economic dispatching method thereof | |
| Lu et al. | Distributed coordination of islanded microgrid clusters using a two-layer intermittent communication network | |
| CN105429185B (en) | A kind of economic load dispatching method with the harmonious property of robust | |
| Liu et al. | Distributed optimal tie-line power flow control for multiple interconnected AC microgrids | |
| CN108494022B (en) | Accurate scheduling method based on distributed power supply in micro-grid | |
| CN110165714B (en) | Micro-grid integrated scheduling and control method based on extreme dynamic programming algorithm and computer readable storage medium | |
| CN113054688B (en) | Renewable energy cluster output distributed coordination control method considering uncertainty | |
| CN110858718A (en) | Alternating current micro-grid distributed event-driven frequency control method considering economy | |
| CN110707758A (en) | Distributed energy storage system scheduling method for improving new energy consumption | |
| CN110599031A (en) | Micro-grid economic dispatching consistency algorithm based on distributed event trigger control | |
| Robbins et al. | Control of distributed energy resources for reactive power support | |
| CN112671034B (en) | Island type alternating current micro-grid frequency recovery and power optimal distribution method based on distributed two-layer control | |
| Wang et al. | Distributed optimization control schemes applied on offshore wind farm active power regulation | |
| CN111160808A (en) | Distributed event-triggered power system economic dispatching method with uncertain parameters | |
| CN110598894A (en) | Data processing method and device for energy Internet and electronic equipment | |
| La Bella et al. | A fully distributed control scheme for power balancing in distribution networks | |
| CN115395647A (en) | Power distribution station cluster cloud edge cooperative regulation and control method and system based on consistency algorithm | |
| Li et al. | Adaptive droop control of a multibus DC microgrid based on consensus algorithm | |
| CN104868488B (en) | A kind of power network energy-storage system is exerted oneself control method and device | |
| CN108459907A (en) | A kind of balancing energy instruction distributing method of more execution units | |
| Tang et al. | Distributed Consensus‐Based Optimization of Multiple Load Aggregators for Secondary Frequency Control | |
| CN112765851A (en) | Consistency distributed vibration control method for solar sailboard of spacecraft | |
| CN111262277A (en) | System and method for optimizing cluster operation of electric energy routers in energy Internet | |
| Ji et al. | Distributed economic dispatch algorithm in smart grid based on event-triggered and fixed-time consensus methods | |
| Capezza et al. | Hierarchical distributed consensus based economic dispatch of distributed energy resources (ders) for networked microgrids |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |