CN117217495A

CN117217495A - Energy scheduling method and device, electronic equipment and storage medium

Info

Publication number: CN117217495A
Application number: CN202311427870.2A
Authority: CN
Inventors: 黄文琦; 梁凌宇; 赵翔宇; 曹尚; 张焕明
Original assignee: Southern Power Grid Digital Grid Research Institute Co Ltd
Current assignee: Southern Power Grid Digital Grid Research Institute Co Ltd
Priority date: 2023-10-30
Filing date: 2023-10-30
Publication date: 2023-12-12

Abstract

The embodiment of the invention discloses an energy scheduling method, an energy scheduling device, electronic equipment and a storage medium. The method comprises the following steps: aiming at power generation equipment to be scheduled and each electric equipment, respectively acquiring a historical power generation time sequence of the power generation equipment and a historical power utilization time sequence of the electric equipment; based on a wavelet recurrent neural network model, performing feature extraction on the historical power generation time sequence to obtain power generation features, and performing feature extraction on the historical power utilization time sequence to obtain power utilization features; based on the cloud distributed graphic processing unit architecture, predicting a future power generation time sequence according to power generation characteristics and predicting a future power utilization time sequence according to power utilization characteristics; and generating an energy scheduling scheme according to the future power generation time sequence and the future power utilization time sequence corresponding to each electric equipment respectively so as to realize energy scheduling between the power generation equipment and each electric equipment. The technical scheme of the embodiment of the invention can ensure the rationality of energy scheduling.

Description

Energy scheduling method and device, electronic equipment and storage medium

Technical Field

The embodiment of the invention relates to the field of smart grids, in particular to an energy scheduling method, an energy scheduling device, electronic equipment and a storage medium.

Background

Currently, renewable energy sources, including photovoltaic and wind, have been used in integrated power generation systems and smart grids for sites with solar and/or wind energy. However, since solar energy and wind energy have intermittent characteristics, the above power generation system cannot provide constant power generation.

In addition, the power consumption of various consumers may fluctuate over time.

Therefore, in the field of smart grids, how to guarantee the rationality of energy scheduling is of great importance.

Disclosure of Invention

The embodiment of the invention provides an energy scheduling method, an energy scheduling device, electronic equipment and a storage medium, so as to ensure the rationality of energy scheduling.

According to an aspect of the present invention, there is provided an energy scheduling method, which may include:

aiming at power generation equipment to be scheduled and each electric equipment in at least one electric equipment responsible for power supply of the power generation equipment, respectively acquiring a historical power generation time sequence representing historical power generation power of the power generation equipment and a historical power utilization time sequence representing historical power utilization power of the electric equipment;

acquiring a trained wavelet recurrent neural network model, performing feature extraction on a historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation features of power generation equipment, and performing feature extraction on a historical power utilization time sequence to obtain power utilization features of electric equipment;

Based on the cloud distributed graphic processing unit architecture, a future power generation time sequence of the power generation equipment for representing future power generation power can be predicted according to the power generation characteristics, and a future power utilization time sequence of the electric equipment for representing future power utilization power can be predicted according to the power utilization characteristics;

based on the cloud distributed graphic processing unit architecture, an energy scheduling scheme is generated according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric device respectively, so that energy scheduling between power generation equipment and the at least one electric device is realized based on the energy scheduling scheme.

According to another aspect of the present invention, there is provided an energy scheduling apparatus, which may include:

the historical electricity utilization time sequence acquisition module is used for respectively acquiring a historical electricity utilization time sequence representing historical electricity generation power of the power generation equipment and a historical electricity utilization time sequence representing historical electricity utilization power of the electric equipment aiming at each electric equipment of power generation equipment to be scheduled and at least one electric equipment for which the power generation equipment is responsible for supplying power;

the power utilization characteristic obtaining module is used for obtaining a trained wavelet recurrent neural network model, carrying out characteristic extraction on the historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation characteristics of power generation equipment, and carrying out characteristic extraction on the historical power utilization time sequence to obtain power utilization characteristics of electric equipment;

The future electricity utilization time sequence prediction module is used for predicting a future electricity generation time sequence of the power generation equipment for representing future electricity generation power based on the cloud distributed graphic processing unit architecture according to the power generation characteristics, and predicting a future electricity utilization time sequence of the electric equipment for representing future electricity utilization power according to the power utilization characteristics;

and the energy scheduling module is used for generating an energy scheduling scheme according to the future power generation time sequence and the future power utilization time sequence corresponding to at least one electric equipment respectively based on the cloud distributed graphic processing unit architecture so as to realize energy scheduling between the power generation equipment and the at least one electric equipment based on the energy scheduling scheme.

According to another aspect of the present invention, there is provided an electronic device, which may include:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,

the memory stores a computer program executable by the at least one processor, the computer program being executable by the at least one processor to cause the at least one processor to implement the energy scheduling method provided by any embodiment of the present invention when executed.

According to another aspect of the present invention, there is provided a computer readable storage medium having stored thereon computer instructions for causing a processor to implement the energy scheduling method provided by any embodiment of the present invention when executed.

According to the technical scheme, for power generation equipment to be scheduled and each electric equipment in at least one electric equipment for which the power generation equipment is responsible for supplying power, a historical power generation time sequence representing historical power generation power of the power generation equipment and a historical power utilization time sequence representing historical power utilization power of the electric equipment are respectively obtained; acquiring a trained WRNN model, performing feature extraction on a historical power generation time sequence based on the WRNN model to obtain power generation features of power generation equipment, and performing feature extraction on a historical power utilization time sequence to obtain power utilization features of electric equipment; based on the cloud distributed GPU architecture, predicting a future power generation time sequence of the power generation equipment for representing future power generation power according to the power generation characteristics, and predicting a future power utilization time sequence of the power utilization equipment for representing future power utilization power according to the power utilization characteristics; then, based on the cloud distributed GPU architecture, an energy scheduling scheme is generated according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric device respectively, so that energy scheduling between the power generation device and the at least one electric device is realized based on the energy scheduling scheme. According to the technical scheme, the WRNN model and the cloud distributed GPU architecture are combined, so that a better energy scheduling scheme can be generated for the power generation equipment and each electric equipment, and the rationality of energy scheduling between the power generation equipment and each electric equipment is effectively guaranteed based on the energy scheduling scheme.

It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention, nor is it intended to be used to limit the scope of the invention. Other features of the present invention will become apparent from the description that follows.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is apparent that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a flowchart of an energy scheduling method according to an embodiment of the present invention;

FIG. 2 is a flow chart of another energy scheduling method provided in accordance with an embodiment of the present invention;

FIG. 3 is a flow chart of yet another energy scheduling method provided in accordance with an embodiment of the present invention;

FIG. 4 is a flow chart of an alternative example of yet another energy scheduling method provided in accordance with an embodiment of the present invention;

fig. 5 is a block diagram of an energy scheduling apparatus according to an embodiment of the present invention;

fig. 6 is a schematic structural diagram of an electronic device implementing an energy scheduling method according to an embodiment of the present invention.

Detailed Description

In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.

It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. The cases of "target", "original", etc. are similar and will not be described in detail herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.

Fig. 1 is a flowchart of an energy scheduling method according to an embodiment of the present invention. The embodiment is applicable to the situation of energy scheduling, in particular to the situation of electric energy scheduling between the power generation equipment and at least one electric equipment. The method can be implemented by the energy scheduling device provided by the embodiment of the invention, the device can be implemented by software and/or hardware, and the device can be integrated on electronic equipment, and the electronic equipment can be various user terminals or servers.

Referring to fig. 1, the method of the embodiment of the present invention specifically includes the following steps:

s110, respectively acquiring a historical power generation time sequence of the power generation equipment representing historical power generation power and a historical power utilization time sequence of the power utilization equipment representing historical power utilization power aiming at the power generation equipment to be scheduled and each power utilization equipment of at least one power utilization equipment for which the power generation equipment is responsible for supplying power.

A power generation device is understood to mean a device, in which the electrical energy it generates is to be scheduled between at least one consumer. The electric device is understood to be a device operated by using electric energy provided by the power generation device, and in practical application, the number of the electric devices may be one, two or more, which is related to the practical situation and is not specifically limited herein.

The method comprises the steps of obtaining historical power generation power of the power generation equipment at each historical power generation time point, and generating a historical power generation time sequence based on each obtained historical power generation power, so that the historical power generation time sequence is obtained. Similarly, for each of the at least one powered device, historical power consumption of the powered device at each historical power consumption time point is obtained, and a historical power consumption time sequence is generated based on each obtained historical power consumption, so that the historical power consumption time sequence is obtained. In practical application, optionally, each historical power generation time point and each historical power utilization time point may be in one-to-one correspondence, or may have a difference, which is not specifically limited herein.

S120, acquiring a trained wavelet recurrent neural network model, performing feature extraction on the historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation features of power generation equipment, and performing feature extraction on the historical power utilization time sequence to obtain power utilization features of electric equipment.

The wavelet recurrent neural network (Wavelet Recurrent Neural Network, WRNN) model is understood, among other things, as a model that has been pre-trained at least for achieving feature extraction.

And acquiring a WRNN model, and extracting features of the historical power generation time sequence based on the WRNN model to obtain power generation features of the power generation equipment, so that redundant data in the historical power generation time sequence is reduced. In the embodiment of the invention, the WRNN model can be used for calculating the data characteristics such as the mean value and/or the variance of each historical power in the historical power generation time sequence, and then the power generation change trend is fitted according to the data characteristics by combining a linear fitting mode or a difference fitting mode and the like, so that the power generation characteristics are obtained.

Similarly, for each electric equipment in at least one electric equipment, based on the WRNN model, the historical electricity utilization time sequence of the electric equipment is subjected to feature extraction to obtain electricity utilization features of the electric equipment, so that redundant data in the historical electricity utilization time sequence are reduced.

In practical application, optionally, in order to ensure the feature extraction effect of the WRNN model, the historical power generation time sequence may be subjected to wavelet decomposition, and then the obtained wavelet decomposition result is input into the WRNN model, so that the power utilization feature of the electric equipment may be obtained according to the output result output by the WRNN model. Similarly, the historical electricity utilization time sequence can be subjected to wavelet decomposition, and then the obtained wavelet decomposition result is input into the WRNN model, so that the electricity utilization characteristics of the electric equipment are obtained according to the output result of the WRNN model. The wavelet decomposition result obtained above can be understood as data that can be directly processed by the WRNN model.

S130, based on the cloud distributed graphic processing unit architecture, predicting a future power generation time sequence of the power generation equipment for representing future power generation power according to the power generation characteristics, and predicting a future power utilization time sequence of the electric equipment for representing future power utilization power according to the power utilization characteristics.

Among them, the cloud distributed graphics processing unit (Graphics Processing Unit, GPU) architecture, which can be understood as an architecture of cloud distributed computing combined with a GPU, makes possible a fast prediction for future power generation time sequences and future power consumption time sequences with large data volumes.

Based on the cloud distributed GPU architecture, future power generation powers of the power generation device at each future power generation time point are predicted according to the power generation characteristics, and a future power generation time series is generated based on these future power generation powers. Similarly, for each of the at least one powered device, based on the cloud-distributed GPU architecture, future power consumption of the powered device at each future power consumption time point can be predicted according to the power consumption characteristics of the powered device, and a future power consumption time sequence of the powered device is generated based on the future power consumption power. In practical applications, it can be understood that each future power generation time point and each future power utilization time point may be in one-to-one correspondence, or may have a difference, which is not specifically limited herein.

And S140, generating an energy scheduling scheme based on the cloud distributed graphic processing unit architecture according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric equipment respectively, so as to realize energy scheduling between the power generation equipment and the at least one electric equipment based on the energy scheduling scheme.

After the future power generation time sequence and the future power utilization time sequence corresponding to each electric equipment respectively are obtained, an energy scheduling scheme can be generated based on the cloud distributed GPU architecture according to the time sequences, the energy scheduling scheme can furthest reduce the influence of power production fluctuation and ensure that the electric equipment meeting variable requirements reliably flows to the electric equipment, so that all energy requirements are met.

Further, energy scheduling between the power generation equipment and each electric equipment is achieved based on the energy scheduling scheme.

An optional technical scheme, the energy scheduling method further includes:

based on the wavelet recurrent neural network model, processing the historical power generation time sequence to obtain each future power generation time point of the power generation equipment for predicting future power generation power;

based on the cloud distributed graphics processing unit architecture, predicting a future generation time sequence of the power generation equipment representing future generation power according to the generation characteristics, comprising:

based on the cloud distributed graphics processing unit architecture, a future power generation time sequence is predicted according to the power generation characteristics and each future power generation time point, wherein the future power generation time sequence represents the future power generation power of the power generation equipment at each future power generation time point.

The future generation time point is understood as a time point of the future generation power of the power generation apparatus to be predicted. And processing the historical power generation time sequence based on the WRNN model to obtain each future power generation time point. Further, with the cloud-distributed GPU architecture, future generation power of the power generation device at each future generation time point can be predicted from the power generation characteristics and each future generation time point, thereby generating a future generation time sequence.

According to the technical scheme, the future generation time points are predicted and then input into the cloud distributed GPU architecture, so that accurate prediction of the future generation time sequence is realized.

In practical application, optionally, the future power utilization time sequence of each electric equipment can be predicted in the same way, and will not be described herein.

Fig. 2 is a flowchart of another energy scheduling method according to an embodiment of the present invention. The present embodiment is optimized based on the above technical solutions. In this embodiment, optionally, based on a cloud distributed graphics processing unit architecture, generating the energy scheduling scheme according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric device respectively may include: generating at least one candidate scheduling scheme according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric equipment respectively based on a cloud distributed graphic processing unit architecture; and acquiring preset power distribution constraints, and determining an energy scheduling scheme meeting the power distribution constraints from at least one candidate scheduling scheme. Wherein, the explanation of the same or corresponding terms as the above embodiments is not repeated herein.

Referring to fig. 2, the method of this embodiment may specifically include the following steps:

s210, respectively acquiring a historical power generation time sequence representing historical power generation power of the power generation equipment and a historical power utilization time sequence representing historical power utilization power of the power utilization equipment aiming at the power generation equipment to be scheduled and each power utilization equipment in at least one power utilization equipment for which the power generation equipment is responsible for supplying power.

S220, acquiring a trained wavelet recurrent neural network model, performing feature extraction on the historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation features of power generation equipment, and performing feature extraction on the historical power utilization time sequence to obtain power utilization features of electric equipment.

S230, based on the cloud distributed graphic processing unit architecture, predicting a future power generation time sequence of the power generation equipment for representing future power generation power according to the power generation characteristics, and predicting a future power utilization time sequence of the power utilization equipment for representing future power utilization power according to the power utilization characteristics.

S240, generating at least one candidate scheduling scheme according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric equipment respectively based on the cloud distributed graphic processing unit architecture.

S250, acquiring preset power distribution constraint, and determining an energy scheduling scheme meeting the power distribution constraint from at least one candidate scheduling scheme so as to realize energy scheduling between the power generation equipment and at least one electric equipment based on the energy scheduling scheme.

The power distribution constraint can be understood as a preset relevant constraint used for representing the condition that the power generation equipment distributes electric energy to each electric equipment. And acquiring the power distribution constraint, and then determining an energy scheduling scheme which can meet the power distribution constraint from at least one candidate scheduling scheme, thereby ensuring the rationality of the energy scheduling scheme of the final application.

According to the technical scheme provided by the embodiment of the invention, the obtained energy scheduling scheme can be ensured to meet the actual electric energy distribution requirement by applying the power distribution constraint, so that the rationality of electric energy scheduling is further ensured.

An optional technical solution, the cloud distributed graphics processing unit architecture includes at least one graphics processing unit thread, at least one candidate scheduling scheme being generated in parallel by different graphics processing unit threads of the at least one graphics processing unit thread;

from the at least one candidate scheduling scheme, determining an energy scheduling scheme that satisfies the power distribution constraint, comprising:

judging whether the candidate scheduling schemes meet the power distribution constraint by utilizing a graphic processing unit thread which generates the candidate scheduling schemes in at least one graphic processing unit thread aiming at each candidate scheduling scheme in the at least one candidate scheduling schemes;

and determining the energy scheduling scheme meeting the power distribution constraint from the at least one candidate scheduling scheme according to the judging result of the graphic processing unit threads respectively corresponding to the at least one candidate scheduling scheme.

In other words, the cloud distributed graphics processing unit architecture comprises at least one GPU thread, different candidate scheduling schemes can be generated based on the GPU threads in parallel, and then whether the generated candidate scheduling schemes meet power distribution constraint is judged by utilizing the GPU threads in parallel, so that the determination efficiency of the energy scheduling scheme is ensured.

Fig. 3 is a flowchart of yet another energy scheduling method provided in an embodiment of the present invention. The present embodiment is optimized based on the above technical solutions. In this embodiment, optionally, for any one of the at least one electric device, in a case where the electric quantity allocated to the electric device by the power generating device does not meet the power demand of the electric device, the electric device obtains the electric quantity from the energy provider; the electric quantity owned by the energy supplier is derived from power generation equipment; then determining an energy scheduling scheme that satisfies the power distribution constraint from the at least one candidate scheduling scheme, comprising: determining at least one intermediate scheduling scheme satisfying the power distribution constraint from the at least one candidate scheduling scheme; determining an energy scheduling scheme from at least one intermediate scheduling scheme based on minimizing the electric quantity obtained by at least one electric equipment from an energy provider; based on the energy scheduling scheme, realizing energy scheduling between the power generation equipment and at least one electric equipment, comprising: based on the energy scheduling scheme, energy scheduling between the power generation equipment and the energy provider and at least one consumer is achieved. Wherein, the explanation of the same or corresponding terms as the above embodiments is not repeated herein.

Referring to fig. 3, the method of this embodiment may specifically include the following steps:

s310, respectively acquiring a historical power generation time sequence representing historical power generation power of the power generation equipment and a historical power utilization time sequence representing historical power utilization power of the power utilization equipment aiming at the power generation equipment to be scheduled and each power utilization equipment in at least one power utilization equipment for which the power generation equipment is responsible for supplying power.

S320, acquiring a trained wavelet recurrent neural network model, performing feature extraction on the historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation features of power generation equipment, and performing feature extraction on the historical power utilization time sequence to obtain power utilization features of electric equipment.

S330, based on the cloud distributed graphic processing unit architecture, predicting a future power generation time sequence of the power generation equipment for representing future power generation power according to the power generation characteristics, and predicting a future power utilization time sequence of the power utilization equipment for representing future power utilization power according to the power utilization characteristics.

S340, generating at least one candidate scheduling scheme according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric equipment respectively based on the cloud distributed graphic processing unit architecture.

S350, acquiring preset power distribution constraints, and determining at least one intermediate scheduling scheme meeting the power distribution constraints from at least one candidate scheduling scheme;

the power generation device is used for generating power, wherein the power generation device is used for generating power, the power supply device is used for supplying power to at least one power consumption device, and the power consumption device is used for supplying power to the power consumption device.

In other words, the remaining energy produced locally is utilized as a further source of commercial energy suppliers, so that the energy suppliers can supply electrical energy to the respective consumers.

On this basis, optionally, the power distribution constraints may include at least one of:

fairness constraints for energy allocation for each consumer and energy provider; the method comprises the steps of,

for any one of the at least one powered device, the difference power of the powered device is less than or equal to the power obtained by the powered device from the power generation device, wherein the difference power is equal to the power obtained by the powered device from the power generation device + the power obtained from the energy supplier-a power tolerance preset for each of the at least one powered device.

S360, determining an energy scheduling scheme from at least one intermediate scheduling scheme based on the basis of minimizing the electric quantity acquired by at least one electric equipment from the energy supplier, so as to realize energy scheduling between the power generation equipment and the energy supplier and at least one electric equipment based on the energy scheduling scheme.

The intermediate scheduling scheme may be understood as a candidate scheduling scheme satisfying the power distribution constraint among the candidate scheduling schemes. The number of intermediate scheduling schemes may be one, two or more, which is relevant to the actual situation and is not specifically limited herein.

Based on the minimization of the electric quantity obtained by each electric device from the energy supplier, the energy scheduling scheme is determined from at least one intermediate scheduling scheme, which is helpful for reducing the electric quantity purchased by each electric device from the energy supplier, thereby ensuring the reliability of energy scheduling.

According to the technical scheme provided by the embodiment of the invention, the energy scheduling scheme is determined by taking the electric quantity obtained by each electric equipment from the energy supplier as the basis, so that the reliability of energy scheduling is ensured.

In order to better understand the above-described respective technical solutions as a whole, an exemplary description thereof is given below in conjunction with specific examples. Exemplary, referring to fig. 4, the specific steps are as follows:

Step 1: collecting measurements for power generation equipment (e.g., power plants, etc.)Historical power generation time series of arrival and historical power utilization time series measured for electric equipment (such as a power utilization building and the like). For general purposes, the above time sequence may be referred to as u (τ _n ) Wherein τ _n Is a discrete time step of the sampled data and may be set to half an hour in practical applications.

Step 2: calculation of u (τ) _n ) And (3) biorthogonal wavelet decomposition of the WRNN model, thereby obtaining an input set of WRNN models. In particular, the biorthogonal wavelet decomposition may be performed by applying wavelet transforms: i-th iterates in this wayFor u (τ) _n ) Generates a set of coefficients d _i (τ _n ) And residual a _i (τ _n ) So as to:

wherein M andu (τ) _n ) Row and column in (a). Here, parentheses are used as parameters of the wavelet transform +.>Brackets of vectors [ d ] _i (τ _n ),a _i (τ _n )]From two coefficients d _i (τ _n ) And a _i (τ _n ) Composition is prepared. In this example, a may be set ₀ (τ _n )＝x(τ _n ) Wherein x (τ) _n ) Can be understood as the time series of the original measurements, whereas u (τ _n ) It can be understood that the time interval of the predetermined period of time is equal to x (τ _n ) And a time sequence with a shorter length is obtained after interception, so that the WRNN model processing is facilitated. Further, the input set may be expressed as +. >Matrix (S)>Time step M horizontal wavelet decomposition, where n-th rows represent time step τ _n As a decomposition:

u(τ _n )＝[d ₁ (τ _n ),d ₂ (τ _n ),…,d _M (τ _n ),a _M (τ _n )]

wherein u (τ) _n ) The wavelet decomposition result is input into the WRNN model. In practical use, u (τ) after wavelet decomposition _n ) Is provided as an input value to the input neurons of the WRNN model proposed by M. Further, for each time step τ _n Predicting future time points τ using WRNN model _n+r Such as future generation time points and future usage time points, where r is the number of future time steps.

Wherein,representing the results after processing by the processing functions in the neural unit.

Step 3: the peak energy sold to the energy suppliers is minimized by maintaining a constant power and smooth variation over time. In particular, a power generation device and a certain number of consumers are taken as examples. Enumerating consumers k e [1, Q using index]N, where Q is a rational number set, and then a special user (k=0) is added, denoted asRepresenting an energy provider who wishes to sell a portion of the electricity to it. For discrete time steps τ _n Each consumer k will be characterized by a power load +.>Therefore, the power balance must be maintained as:

Wherein,representing the amount of electricity allocated to the consumer by the power generation device, +.>Representing the amount of electricity from the energy provider. For each time step τ _n The power tolerance delta can be negotiated with each electric equipment ^k (τ _n ) So as to:

wherein,the difference electric quantity is the above.

And presumes that:

wherein P is _G (τ _n ) Is the total power generation.

In a real scenario, it is often necessary to consider fairness constraints for energy allocation among consumers, so that the ratio of total energy allocation shares reaches ρ granted by contract _k Can impose:

wherein t is a continuous variable, Δt is contractual andthe amount of time is taken into account to assign a mathematical meaning k to each integral Represents->

Within a given definition and allocation constraint, a number of different distribution options are conceivable and correspond to one of the settings:

this determines the power allocated to the different consumers. In this step, attention may be paid to a smooth change in the power granted to the provider, which is represented in mathematical form as the consumer number 0. Power generation equipment selling part power generation amount P _G (τ _n ) To energy suppliers asFurthermore, due to the nature of the energy contract, stability +_ can be considered at first >Thus, at each time step τ _n Find the best setting S _* (τ _n ) (i.e., the time series predicted by the WRNN model), and therefore:

wherein,representing the first derivation of a distributed power supply (i.e. powered device) at a certain time stepTo the energy supplier for setting S (τ _n ) Module τ _n 。

Step 4: the input data for the cloud-distributed GPU architecture may be P consisting of power generation characteristics and a future power generation time series _G (τ _n ). Each time step requires a candidate scheduling scheme that satisfies all of the distribution constraints, so GPU threads are organized into blocks and dedicated to each time step. Thus, each thread in each block proposes a different candidate scheduling scheme, which may consist of a setting of the powered device. Of course, if the possible settings do not meet the given power distribution constraints, such settings are deleted. At the end of this step, optimally set S _* (τ _n ) Will be selected. By collecting all the results, a future power generation time series S to the predicted optimal setting can be obtained _* (τ _n )。

Step 5: the prediction data is fed into a cloud-based virtual machine equipped with a GPU to construct a simulated scene accordingly. Each GPU thread is assigned a subset of the generator and load node combination, thereby calculating an optimal distribution of such subsets. Because of the nature of the problem, a thorough simulation can be performed using the GPU thread, i.e. taking into account all possible combinations. Of course, the greater the number of generators and load nodes, the greater the number of combinations possible, revealing the usefulness of the cloud-GPU hybrid approach. When the optimal solution is found by a greedy exclusion algorithm, the optimal solution is given as an output for early reconfiguration of the energy schedule to minimize the overall cost of wasting energy overproduction and/or external purchases due to local excess power load.

Step 6: in order to evaluate cloud-based service performance and check scalability, the amount of data analyzed is greatly increased by connecting more power generation devices and powered devices. The proposed processing time of the distributed component is measured. By increasing the number of powered devices, the scheduling configuration increases dramatically. The arrangement may be referred to herein as a possible configuration of a power plant with different consumers and energy suppliers. Considering the multiplicity n of consumers, finding which k < consumers must meet the distribution constraints at a given time step may require checking a large number of possible settings n can give energy. Then, the combination k powered devices (satisfying) increases very rapidly for the powered devices available above n.

The above examples present a cloud-distributed GPU architecture, in combination with soft computing techniques for energy scheduling management, aimed at ensuring reliable energy flows when using energy featuring fluctuating and variable energy demands. The above examples may be implemented in a real-time processing manner. The above example enables early prediction of future generated power and future used power, and prediction of future generated time series and future used time series by correctly modeling the phenomenon. Furthermore, by means of the GPU and the prediction data, all possible candidate scheduling schemes of the consumer can be simulated. Then, an optimal energy scheduling scheme is selected to be applied that minimizes the underloaded energy, whereby the reliability of the power plant can be improved by reducing the amount of electricity purchased from the energy supplier.

Fig. 5 is a block diagram of an energy scheduling apparatus according to an embodiment of the present invention, where the apparatus is configured to execute the energy scheduling method according to any of the foregoing embodiments. The device and the energy scheduling method of each embodiment belong to the same invention conception, and reference can be made to the embodiment of the energy scheduling method for details which are not described in detail in the embodiment of the energy scheduling device. Referring to fig. 5, the apparatus specifically includes: a historical electricity usage time series acquisition module 410, an electricity usage characteristics acquisition module 420, a future electricity usage time series prediction module 430, and an energy scheduling module 440.

The historical electricity utilization time sequence obtaining module 410 is configured to obtain, for each of the power generation device to be scheduled and at least one electric device for which the power generation device is responsible for supplying power, a historical electricity generation time sequence representing historical electricity generation power of the power generation device and a historical electricity utilization time sequence representing historical electricity utilization power of the electric device, respectively;

the electricity utilization characteristic obtaining module 420 is configured to obtain a trained wavelet recurrent neural network model, perform characteristic extraction on a historical electricity generation time sequence based on the wavelet recurrent neural network model to obtain electricity generation characteristics of the electricity generation equipment, and perform characteristic extraction on the historical electricity utilization time sequence to obtain electricity utilization characteristics of the electric equipment;

A future electricity consumption time sequence prediction module 430, configured to predict a future electricity generation time sequence of the electricity generation device, which characterizes future electricity generation power, based on the cloud distributed graphics processing unit architecture, according to the electricity generation characteristics, and predict a future electricity consumption time sequence of the electricity consumption device, which characterizes future electricity consumption power, according to the electricity consumption characteristics;

the energy scheduling module 440 is configured to generate an energy scheduling scheme according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric device respectively based on the cloud distributed graphics processing unit architecture, so as to implement energy scheduling between the power generation device and the at least one electric device based on the energy scheduling scheme.

Optionally, the energy scheduling device further includes:

the future generation time point prediction module is used for processing the historical generation time sequence based on the wavelet recurrent neural network model to obtain each future generation time point of the future generation power to be predicted of the power generation equipment;

the future electricity utilization time series prediction module 430 includes:

and the future power generation time sequence prediction unit is used for predicting a future power generation time sequence according to the power generation characteristics and each future power generation time point based on the cloud distributed graphic processing unit architecture, wherein the future power generation time sequence represents the future power generation power of the power generation equipment at each future power generation time point.

Optionally, the energy scheduling module 440 includes:

the candidate scheduling scheme generation sub-module is used for generating at least one candidate scheduling scheme according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric equipment respectively based on a cloud distributed graphic processing unit architecture;

the energy scheduling scheme determining submodule is used for acquiring preset power distribution constraint and determining an energy scheduling scheme meeting the power distribution constraint from at least one candidate scheduling scheme.

On this basis, an optional, cloud-distributed graphics processing unit architecture includes at least one graphics processing unit thread, at least one candidate scheduling scheme being generated in parallel by different ones of the at least one graphics processing unit thread;

an energy scheduling scheme determination sub-module comprising:

the power distribution constraint judging unit is used for judging whether the candidate scheduling schemes meet the power distribution constraint by utilizing the graphic processing unit threads which generate the candidate scheduling schemes in the at least one graphic processing unit threads for each candidate scheduling scheme in the at least one candidate scheduling schemes;

and the energy scheduling scheme determining unit is used for determining the energy scheduling scheme meeting the power distribution constraint from at least one candidate scheduling scheme according to the judging result of the graphic processing unit threads corresponding to the at least one candidate scheduling scheme respectively.

Alternatively, for any one of the at least one electric device, in the case that the electric quantity allocated to the electric device by the power generation device does not meet the electric demand of the electric device, the electric device obtains the electric quantity from the energy supplier, and the electric quantity owned by the energy supplier is derived from the power generation device;

an energy scheduling scheme determination sub-module comprising:

an intermediate scheduling scheme determining unit configured to determine at least one intermediate scheduling scheme satisfying a power distribution constraint from among the at least one candidate scheduling schemes;

the energy scheduling scheme determining unit is used for determining an energy scheduling scheme from at least one intermediate scheduling scheme on the basis of minimizing the electric quantity obtained by at least one electric equipment from an energy provider;

the energy scheduling module 440 further includes:

and the energy scheduling sub-module is used for realizing energy scheduling between the power generation equipment, the energy supplier and at least one electric equipment based on an energy scheduling scheme.

On this basis, optionally, the power distribution constraints include at least one of:

for any one of the at least one electric device, the difference electric quantity of the electric device is smaller than or equal to the electric quantity obtained by the electric device from the power generation device, wherein the difference electric quantity is equal to the electric quantity obtained by the electric device from the power generation device+the electric quantity obtained from the energy supplier-the electric quantity tolerance preset for each electric device in the at least one electric device.

Optionally, the electricity utilization feature obtaining module 420 includes:

the wavelet decomposition result input unit is used for carrying out wavelet decomposition on the historical power generation time sequence and inputting the obtained wavelet decomposition result into the wavelet recurrent neural network model;

and the electricity utilization characteristic obtaining unit is used for obtaining the electricity utilization characteristics of the electric equipment according to the output result output by the wavelet recurrent neural network model.

According to the energy scheduling device provided by the embodiment of the invention, the historical power generation time sequence of the power generation equipment and the historical power utilization time sequence of the power utilization equipment, which characterizes the historical power generation power, are respectively acquired by the historical power utilization time sequence acquisition module aiming at the power generation equipment to be scheduled and each power utilization equipment of at least one power utilization equipment for which the power generation equipment is responsible for supplying power; the power utilization characteristic obtaining module is used for obtaining a pre-trained WRNN model, carrying out characteristic extraction on a historical power generation time sequence based on the WRNN model to obtain power generation characteristics of power generation equipment, and carrying out characteristic extraction on the historical power utilization time sequence to obtain power utilization characteristics of electric equipment; predicting a future power generation time sequence representing the future power generation power of the power generation equipment according to the power generation characteristics and predicting a future power utilization time sequence representing the future power utilization power of the electric equipment according to the power utilization characteristics by a future power utilization time sequence prediction module based on a cloud distributed GPU architecture; then, an energy scheduling scheme is generated based on the cloud distributed GPU architecture according to a future power generation time sequence and a future power utilization time sequence corresponding to at least one electric device respectively, so that energy scheduling between the power generation device and the at least one electric device is realized based on the energy scheduling scheme. According to the device, by combining the WRNN model and the cloud distributed GPU architecture, a better energy scheduling scheme can be generated for the power generation equipment and each electric equipment, and the rationality of energy scheduling between the power generation equipment and each electric equipment is effectively ensured based on the better energy scheduling scheme.

The energy scheduling device provided by the embodiment of the invention can execute the energy scheduling method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

It should be noted that, in the embodiment of the energy scheduling device, each unit and module included are only divided according to the functional logic, but not limited to the above-mentioned division, so long as the corresponding functions can be implemented; in addition, the specific names of the functional units are also only for distinguishing from each other, and are not used to limit the protection scope of the present invention.

Fig. 6 shows a schematic diagram of the structure of an electronic device 10 that may be used to implement an embodiment of the invention. Electronic devices are intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.

As shown in fig. 6, the electronic device 10 includes at least one processor 11, and a memory, such as a Read Only Memory (ROM) 12, a Random Access Memory (RAM) 13, etc., communicatively connected to the at least one processor 11, in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various appropriate actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.

Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.

The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as the energy scheduling method.

In some embodiments, the energy scheduling method may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the energy scheduling method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the energy scheduling method in any other suitable way (e.g., by means of firmware).

Various implementations of the systems and techniques described here above may be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.

A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.

In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.

The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.

The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.

It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.

The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.

Claims

1. An energy scheduling method, comprising:

aiming at power generation equipment to be scheduled and each electric equipment in at least one electric equipment for which the power generation equipment is responsible for supplying power, respectively acquiring a historical power generation time sequence representing historical power generation power of the power generation equipment and a historical power utilization time sequence representing historical power utilization power of the electric equipment;

acquiring a trained wavelet recurrent neural network model, extracting features of the historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation features of the power generation equipment, and extracting features of the historical power utilization time sequence to obtain power utilization features of the electric equipment;

Based on a cloud distributed graphic processing unit architecture, predicting a future power generation time sequence of the power generation equipment representing future power generation power according to the power generation characteristics, and predicting a future power utilization time sequence of the electric equipment representing future power utilization power according to the power utilization characteristics;

and generating an energy scheduling scheme based on the cloud distributed graphic processing unit architecture according to the future power generation time sequence and the future power utilization time sequence corresponding to the at least one electric device respectively, so as to realize energy scheduling between the power generation equipment and the at least one electric device based on the energy scheduling scheme.

2. The method as recited in claim 1, further comprising:

processing the historical power generation time sequence based on the wavelet recurrent neural network model to obtain each future power generation time point of the power generation equipment for future power generation to be predicted;

the cloud-based distributed graphics processing unit architecture predicts a future generation time sequence of the power generation device, which characterizes future generation power, according to the generation characteristics, and comprises:

based on a cloud-distributed graphics processing unit architecture, a future power generation time series is predicted from the power generation characteristics and the future power generation time points, wherein the future power generation time series characterizes future power generation power of the power generation device at the future power generation time points.

3. The method of claim 1, wherein the generating, based on the cloud distributed graphics processing unit architecture, an energy scheduling scheme according to the future power generation time sequence and the future power usage time sequence respectively corresponding to the at least one powered device comprises:

generating at least one candidate scheduling scheme according to the future power generation time sequence and the future power utilization time sequence corresponding to the at least one electric equipment respectively based on the cloud distributed graphic processing unit architecture;

and acquiring preset power distribution constraints, and determining an energy scheduling scheme meeting the power distribution constraints from the at least one candidate scheduling scheme.

4. The method of claim 3, wherein the cloud distributed graphics processing unit architecture comprises at least one graphics processing unit thread, the at least one candidate scheduling scheme being generated in parallel by different ones of the at least one graphics processing unit thread;

the determining an energy scheduling scheme satisfying the power distribution constraint from the at least one candidate scheduling scheme includes:

for each candidate scheduling scheme in the at least one candidate scheduling scheme, judging whether the candidate scheduling scheme meets the power distribution constraint by utilizing a graphic processing unit thread which generates the candidate scheduling scheme in the at least one graphic processing unit thread;

And determining an energy scheduling scheme meeting the power distribution constraint from the at least one candidate scheduling scheme according to the judging result of the graphic processing unit threads respectively corresponding to the at least one candidate scheduling scheme.

5. The method of claim 3, wherein for any one of the at least one powered device, the powered device obtains power from an energy provider if the power demand of the powered device is not satisfied by the power allocated to the powered device by the power generation device;

the electric quantity owned by the energy supplier is derived from the power generation equipment;

determining at least one intermediate scheduling scheme satisfying the power distribution constraint from the at least one candidate scheduling scheme;

determining an energy scheduling scheme from the at least one intermediate scheduling scheme based on minimizing the electric quantity obtained by the at least one electric equipment from the energy provider;

the energy scheduling scheme based on the energy scheduling scheme realizes energy scheduling between the power generation equipment and the at least one electric equipment, and comprises the following steps:

And based on the energy scheduling scheme, energy scheduling between the power generation equipment, the energy provider and the at least one electric equipment is realized.

6. The method of claim 5, wherein the power distribution constraints comprise at least one of:

fairness constraints for energy allocation for said each consumer and said energy provider; the method comprises the steps of,

7. The method according to claim 1, wherein the feature extraction of the historical power generation time series based on the wavelet recurrent neural network model to obtain power generation features of the power generation equipment comprises:

performing wavelet decomposition on the historical power generation time sequence, and inputting an obtained wavelet decomposition result into the wavelet recurrent neural network model;

And obtaining the electricity utilization characteristics of the electric equipment according to the output result output by the wavelet recurrent neural network model.

8. An energy scheduling apparatus, comprising:

the historical electricity utilization time sequence acquisition module is used for respectively acquiring a historical electricity utilization time sequence of the power generation equipment representing the historical electricity generation power and a historical electricity utilization time sequence of the electric equipment representing the historical electricity utilization power aiming at each electric equipment in the power generation equipment to be scheduled and at least one electric equipment for which the power generation equipment is responsible for supplying power;

the power utilization characteristic obtaining module is used for obtaining a trained wavelet recurrent neural network model, carrying out characteristic extraction on the historical power generation time sequence based on the wavelet recurrent neural network model to obtain power generation characteristics of the power generation equipment, and carrying out characteristic extraction on the historical power utilization time sequence to obtain power utilization characteristics of the electric equipment;

the future electricity utilization time sequence prediction module is used for predicting a future electricity generation time sequence of the power generation equipment, representing future electricity generation power, based on a cloud distributed graphic processing unit architecture according to the power generation characteristics, and predicting a future electricity utilization time sequence of the electric equipment, representing future electricity utilization power, according to the electricity utilization characteristics;

And the energy scheduling module is used for generating an energy scheduling scheme according to the future power generation time sequence and the future power utilization time sequence corresponding to the at least one electric equipment respectively based on the cloud distributed graphic processing unit architecture so as to realize energy scheduling between the power generation equipment and the at least one electric equipment based on the energy scheduling scheme.

9. An electronic device, comprising:

at least one processor; and

a memory communicatively coupled to the at least one processor; wherein,

the memory stores a computer program executable by the at least one processor to cause the at least one processor to perform the energy scheduling method of any one of claims 1-7.

10. A computer readable storage medium storing computer instructions for causing a processor to perform the energy scheduling method of any one of claims 1-7.