CN108830431A - A kind of Electricity price forecasting solution and relevant apparatus based on whale optimization algorithm - Google Patents
A kind of Electricity price forecasting solution and relevant apparatus based on whale optimization algorithm Download PDFInfo
- Publication number
- CN108830431A CN108830431A CN201810877725.7A CN201810877725A CN108830431A CN 108830431 A CN108830431 A CN 108830431A CN 201810877725 A CN201810877725 A CN 201810877725A CN 108830431 A CN108830431 A CN 108830431A
- Authority
- CN
- China
- Prior art keywords
- electricity price
- population
- prediction
- neural network
- elite
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000005611 electricity Effects 0.000 title claims abstract description 119
- 241000283153 Cetacea Species 0.000 title claims abstract description 113
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 100
- 238000005457 optimization Methods 0.000 title claims abstract description 94
- 238000013528 artificial neural network Methods 0.000 claims abstract description 80
- 238000012545 processing Methods 0.000 claims abstract description 60
- 238000011160 research Methods 0.000 claims abstract description 48
- 230000014759 maintenance of location Effects 0.000 claims abstract description 36
- 238000012549 training Methods 0.000 claims abstract description 15
- 238000003860 storage Methods 0.000 claims abstract description 14
- 238000000034 method Methods 0.000 claims description 37
- 230000008569 process Effects 0.000 claims description 28
- 238000012216 screening Methods 0.000 claims description 12
- 238000004590 computer program Methods 0.000 claims description 9
- 238000000354 decomposition reaction Methods 0.000 claims description 9
- 210000004218 nerve net Anatomy 0.000 claims description 7
- 230000001537 neural effect Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- 235000013399 edible fruits Nutrition 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims 1
- 238000012360 testing method Methods 0.000 claims 1
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 239000002245 particle Substances 0.000 description 7
- 241001456553 Chanodichthys dabryi Species 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 230000002068 genetic effect Effects 0.000 description 4
- 238000013178 mathematical model Methods 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 4
- 230000006870 function Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 241000251468 Actinopterygii Species 0.000 description 1
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000008602 contraction Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- 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
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0283—Price estimation or determination
-
- 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—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Tourism & Hospitality (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
This application discloses a kind of Electricity price forecasting solutions based on whale optimization algorithm, including：Optimization processing is iterated to the result of whale optimization algorithm according to elite retention strategy and crossover operation, obtains the weight and threshold value of neural network；Parameter adjustment is carried out to the initial neural network of building according to the weight and threshold value, prediction neural network is obtained and original electricity price sequence is pre-processed, prediction processing is carried out to pre-processed results using prediction neural network, obtains Research on electricity price prediction result.Whale optimization algorithm is optimized by elite retention strategy and crossover operation, solve the problems, such as that the whale optimization algorithm later period ability of searching optimum of search is poor, and then solve the problems, such as that parameter training existing for neural network is insufficient, neural network is improved for the accuracy rate and availability of Research on electricity price prediction.Disclosed herein as well is a kind of Research on electricity price prediction device, server and computer readable storage mediums based on whale optimization algorithm, have the above beneficial effect.
Description
Technical field
This application involves field of computer technology, in particular to a kind of Electricity price forecasting solution based on whale optimization algorithm,
Research on electricity price prediction device, server and computer readable storage medium.
Background technique
With the continuous development of computer technology, real-life quantifiable data is adopted by computer technology
Collection, then the developing direction of data is predicted by the rule that machine learning obtains, can greatly it optimize in daily production
In configuration to resource.Wherein, in energy internet and electricity market, the optimization that accurate Research on electricity price prediction is conducive to the energy is matched
It sets, reasonably bidding strategy and electricity consumption resident customize reasonably Transaction algorithm for electricity provider customization.
Further, excellent using swarm intelligence algorithm in the prior art in order to cope with the non-linear and randomness of electricity price sequence
Change neural network to predict electricity price, specific such as genetic algorithm, particle swarm algorithm.But genetic algorithm is in actual use
Runing time is longer, and local search ability is poor；Particle swarm algorithm keeps it complete in the later period because of the reason of population location updating mechanism
Office's search capability is poor, reduces neural network for the accuracy rate and availability of Research on electricity price prediction.
Therefore, how to improve neural network is those skilled in the art's focal point to the accuracy rate and validity of electricity price sequence
Problem.
Summary of the invention
The purpose of the application is to provide a kind of Electricity price forecasting solution based on whale optimization algorithm, Research on electricity price prediction device, clothes
Business device and computer readable storage medium, optimize whale optimization algorithm by elite retention strategy and crossover operation,
Search stability and optimizing ability of the whale group algorithm when finding the weight and threshold values of neural network are improved, and then solves mind
Through the insufficient problem of parameter training existing for network, neural network is improved for the accuracy rate and availability of Research on electricity price prediction.
In order to solve the above technical problems, the application provides a kind of Electricity price forecasting solution based on whale optimization algorithm, including：
Optimization processing is iterated to the result of whale optimization algorithm according to elite retention strategy and crossover operation, obtains mind
Weight and threshold value through network；Parameter adjustment is carried out to the initial neural network of building according to the weight and threshold value, is obtained pre-
Survey neural network；
Original electricity price sequence is pre-processed, pre-processed results are carried out at prediction using the prediction neural network
Reason, obtains Research on electricity price prediction result.
Optionally, the result of whale optimization algorithm is iterated at optimization according to elite retention strategy and crossover operation
Reason, obtains the weight and threshold value of neural network, including：
Random initializtion processing is carried out to whale optimization algorithm, obtains initialization population；
According to whale optimization algorithm to the initialization population carry out whale group update processing, by obtained progeny population with
Initialization population merges into mixed population, carries out Screening Treatment to the mixed population according to the elite retention strategy and obtains essence
English population；
Crossover operation is executed to the elite population, obtained intersection progeny population and the elite population are merged into friendship
Mixed population is pitched, Screening Treatment is carried out to the cross-mixing population according to the elite retention strategy, obtains intersecting elite kind
Group；
The intersection elite population as the initialization population and is executed into corresponding processing, is reached until executing number
Maximum number of iterations, using the position of optimum individual in the intersection elite population as the weight of neural network and threshold value.
Optionally, original electricity price sequence is pre-processed, pre-processed results is carried out using the prediction neural network
Prediction processing, obtains Research on electricity price prediction as a result, including：
Resolution process is carried out to the original electricity price sequence and obtains multiple subsequences；
Prediction processing is carried out to all subsequences using the prediction neural network, obtains multiple subsequence predictions
Value；
All subsequence predicted values are overlapped to obtain Research on electricity price prediction result.
Optionally, resolution process is carried out to the original electricity price sequence and obtains multiple subsequences, including：
Rule of thumb pattern decomposition algorithm carries out resolution process to the original electricity price sequence, obtains the multiple sub- sequence
Column.
The application also provides a kind of Research on electricity price prediction device based on whale optimization algorithm, including：
Neural metwork training module, for according to elite retention strategy and crossover operation to the result of whale optimization algorithm into
Row iteration optimization processing obtains the weight and threshold value of neural network；According to the weight and threshold value to the initial nerve net of building
Network carries out parameter adjustment, obtains prediction neural network；
Research on electricity price prediction module, for being pre-processed to original electricity price sequence, using the prediction neural network to pre- place
Reason result carries out prediction processing, obtains Research on electricity price prediction result.
Optionally, the neural metwork training module, including iterative optimization unit and parameter adjustment unit；
Wherein, the iterative optimization unit includes：
Initialization process subelement obtains initialization population for carrying out random initializtion processing to whale optimization algorithm；
Whale group location updating subelement, for carrying out whale group more to the initialization population according to whale optimization algorithm
Obtained progeny population and initialization population are merged into mixed population, according to the elite retention strategy to described by new processing
Mixed population carries out Screening Treatment and obtains elite population；
Population cross processing subelement, for executing crossover operation, the intersection filial generation kind that will be obtained to the elite population
Group merges into cross-mixing population with the elite population, is carried out according to the elite retention strategy to the cross-mixing population
Screening Treatment obtains intersecting elite population；
Iteration result obtains subelement, for as the initialization population and executing corresponding the intersection elite population
Processing, reach maximum number of iterations until executing number, using the position of optimum individual in the intersection elite population as refreshing
Weight and threshold value through network.
Optionally, the Research on electricity price prediction module, including：
Pretreatment unit obtains multiple subsequences for carrying out resolution process to the original electricity price sequence；
Predicting unit obtains more for carrying out prediction processing to all subsequences using the prediction neural network
A sub- sequence prediction value；
As a result acquiring unit, for being overlapped to obtain Research on electricity price prediction result by all subsequence predicted values.
Optionally, the pretreatment unit be specifically used for rule of thumb pattern decomposition algorithm to the original electricity price sequence into
Row resolution process obtains the multiple subsequence.
The application also provides a kind of server, including：
Memory, for storing computer program；
Processor, the step of Electricity price forecasting solution as described above is realized when for executing the computer program.
The application also provides a kind of computer readable storage medium, and calculating is stored on the computer readable storage medium
The step of machine program, the computer program realizes Electricity price forecasting solution as described above when being executed by processor.
A kind of Electricity price forecasting solution based on whale optimization algorithm provided herein, including：Retain plan according to elite
Slightly and crossover operation is iterated optimization processing to the result of whale optimization algorithm, obtains the weight and threshold value of neural network；Root
Parameter adjustment is carried out to the initial neural network of building according to the weight and threshold value, obtains prediction neural network；To original electricity price
Sequence is pre-processed, and is carried out prediction processing to pre-processed results using the prediction neural network, is obtained Research on electricity price prediction result.
Whale optimization algorithm is optimized by elite retention strategy and crossover operation, by optimum results to initial mind
Through network be trained processing improve whale group algorithm find neural network weight and threshold values when search stability and
Optimizing ability, and then solve the problems, such as that parameter training existing for neural network is insufficient, neural network is improved for Research on electricity price prediction
Accuracy rate and availability.
The application also provides a kind of Research on electricity price prediction device based on whale optimization algorithm, server and computer-readable deposits
Storage media has the above beneficial effect, and this will not be repeated here.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of application for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is a kind of flow chart of the Electricity price forecasting solution based on whale optimization algorithm provided by the embodiment of the present application；
Fig. 2 is the flow chart of the iteration optimization processing of Electricity price forecasting solution provided by the embodiment of the present application；
Fig. 3 is the flow chart of the prediction processing of Electricity price forecasting solution provided by the embodiment of the present application；
Fig. 4 is a kind of structural representation of the Research on electricity price prediction device based on whale optimization algorithm provided by the embodiment of the present application
Figure.
Specific embodiment
The core of the application is to provide a kind of Electricity price forecasting solution based on whale optimization algorithm, Research on electricity price prediction device, clothes
Business device and computer readable storage medium, optimize whale optimization algorithm by elite retention strategy and crossover operation,
Search stability and optimizing ability of the whale group algorithm when finding the weight and threshold values of neural network are improved, and then solves mind
Through the insufficient problem of parameter training existing for network, neural network is improved for the accuracy rate and availability of Research on electricity price prediction.
To keep the purposes, technical schemes and advantages of the embodiment of the present application clearer, below in conjunction with the embodiment of the present application
In attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is
Some embodiments of the present application, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall in the protection scope of this application.
In the prior art in order to cope with the non-linear and randomness of electricity price sequence, using swarm intelligence algorithm optimization neural network
Electricity price is predicted, specific such as genetic algorithm, particle swarm algorithm.But genetic algorithm in actual use runing time compared with
Long, local search ability is poor；Particle swarm algorithm makes it in later period ability of searching optimum because of the reason of population location updating mechanism
It is poor, reduce the accuracy rate and availability for Research on electricity price prediction of neural network.
Therefore, the present embodiment provides a kind of Electricity price forecasting solution based on whale optimization algorithm, pass through elite retention strategy
Whale optimization algorithm is optimized with crossover operation, improves whale group algorithm when finding the weight and threshold values of neural network
Search stability and optimizing ability, and then solve the problems, such as that parameter training existing for neural network is insufficient, improve nerve net
Accuracy rate and availability of the network for Research on electricity price prediction.
Specifically, referring to FIG. 1, Fig. 1 is a kind of electricity price based on whale optimization algorithm provided by the embodiment of the present application
The flow chart of prediction technique.
This method may include：
S101 is iterated optimization processing to the result of whale optimization algorithm according to elite retention strategy and crossover operation,
Obtain the weight and threshold value of neural network；Parameter adjustment is carried out to the initial neural network of building according to weight and threshold value, is obtained
Prediction neural network；
This step is intended to carry out the weight and threshold value that neural network is calculated by whale optimization algorithm, that is, passes through
The parameter of whale optimization algorithm optimization neural network, to adjust the parameter of neural network, the prediction nerve net that can be used
Network.The training process of namely general neural network.Optimal whale is usually directly calculated by whale optimization algorithm
Individual realizes the optimization to neural network using the location parameter of whale individual as the weight of neural network and threshold value.But
General whale optimization algorithm is easily trapped into local optimum in the later period of iterative calculation, so that convergence rate is slower.
Therefore, in order to avoid local optimum problem, elite retention strategy and friendship are added to whale optimization algorithm in this step
Fork operation improves ability of searching optimum of the whale optimization algorithm when calculating, avoids falling into local optimum, find optimal nerve
The weight and threshold value of network.
Wherein, elite retention strategy is to retain the individual of the elite in whale optimization algorithm, that is, reservation fitness
Relatively good individual, then cross and variation is carried out to elite individual by crossover operation and obtains random individual, increase whale optimization
The ability of searching optimum of algorithm.This step constantly can carry out elite to obtained progeny population by way of iteration and retain plan
Elite individual is slightly filtered out, then crossover operation is carried out to elite individual and obtains Variant progeny population, by Variant progeny group and essence
English individual merges the progeny population started as iteration, continues to execute iterative operation until reaching default the number of iterations.?
Weight and threshold value of the optimal individual as neural network are selected in whale population after reaching the number of iterations, improve whale optimization
The ability of searching optimum of algorithm.
S102 pre-processes original electricity price sequence, is carried out at prediction using prediction neural network to pre-processed results
Reason, obtains Research on electricity price prediction result.
On the basis of step S101, after this step is intended to handle original electricity price sequence, by predicting nerve net
Network carries out prediction processing to original electricity price sequence, obtains Research on electricity price prediction result.Wherein, the pre- place original electricity price sequence carried out
Reason is usually the processing carried out data in format, form or expression way, and original electricity price sequence can be made to be more suitable for passing through
Prediction neural network carries out prediction processing, obtains Research on electricity price prediction result.
To sum up, the present embodiment optimizes whale optimization algorithm by elite retention strategy and crossover operation, improves
Search stability and optimizing ability of the whale group algorithm when finding the weight and threshold values of neural network, and then solve neural network
The insufficient problem of existing parameter training improves neural network for the accuracy rate and availability of Research on electricity price prediction.
Based on above embodiments, the present embodiment is illustrated mainly for how being iterated optimization processing and do one,
He is partially substantially the same with a upper embodiment, and same section can refer to a upper embodiment, and this will not be repeated here.
Referring to FIG. 2, Fig. 2 is the process of the iteration optimization processing of Electricity price forecasting solution provided by the embodiment of the present application
Figure.
The processing may include：
S201 carries out random initializtion processing to whale optimization algorithm, obtains initialization population；
This step is intended to carry out whale optimization algorithm initialization process, that is, initial ginseng is arranged to whale optimization algorithm
Number, and initialization is carried out by the whale optimization algorithm after setting initiation parameter and calculates whale population, that is, initialization kind
Group.Initialization population is random at this time.
S202 carries out whale group update processing, the progeny population that will be obtained to initialization population according to whale optimization algorithm
Mixed population is merged into initialization population, Screening Treatment is carried out to mixed population according to elite retention strategy and obtains elite kind
Group；
On the basis of step S201, whale group population and initialization kind that this step is intended to update whale optimization algorithm
Group selects elite individual according to elite retention strategy.
Wherein, elite retention strategy is that the optimum individual of current group in order to prevent is lost in the next generation.The strategy
Thought to be advantage individual (namely elite individual) that group is occurred so far during evolution intersect and straight without pairing
It connects and copies in the next generation.In this step it is exactly to merge progeny population and initialization population, retains further according to elite
Policy Filtering goes out elite population.
Wherein, it is to be screened according to the fitness of whale individual, that is, calculate kind that elite retention strategy, which carries out screening,
The fitness of each elite individual, the highest individual of the fitness of preset quantity is filtered out in group.For example, by every in the population
Individual calculates corresponding fitness, is ranked up on this basis to each individual according to fitness, selects first 10
Body is as elite individual.
S203 executes crossover operation to elite population, obtained intersection progeny population is merged into elite population and is intersected
Mixed population carries out Screening Treatment to cross-mixing population according to elite retention strategy, obtains intersecting elite population；
On the basis of step S202, this step is intended to carry out elite population cross compile operation to improve population result
Ability of searching optimum, that is, degree of randomization.
Wherein, when the crossover operation of progress, the new random individual of cross occurrence can be carried out according to population at individual, improves kind
The ability of searching optimum of group.But general crossover operation will lead to current group, i.e., the optimum individual in elite group is under
It is lost in generation group, influences the convergence capabilities of optimization algorithm.Therefore in this step, by elite population and intersect filial generation kind
Group is mixed, and is continuing to screen according to elite retention strategy to cross-mixing population, to guarantee that optimum individual is not lost.?
Exactly in the case where improving the ability of searching optimum of optimization algorithm, it ensure that the convergence capabilities of algorithm, it is available optimal
Individual.
S204 will intersect elite population as initialization population and execute corresponding processing, reaches most until executing number
Big the number of iterations will intersect the position of optimum individual in elite population as the weight of neural network and threshold value.
On the basis of step S202 and step S203, intersection elite population that this step is intended to obtain S203 as
Initialization population and directly execution S202 and S203 in S202.Namely execute S202 and S203 loop iteration, until circulation
The number of execution reaches maximum number of iterations, using intersect elite population in optimum individual position as the weight of neural network with
Threshold value.
Wherein, the present embodiment can be implemented in specific application environment by way of circulation plus judgement, each time
All add an iteration number in circulation, and judge whether the number of iterations is greater than maximum number of iterations, changes when less than or equal to maximum
When generation number, circulation is continued to execute, when being greater than maximum number of iterations, stops loop iteration, executes subsequent step, that is, will
Intersect weight and threshold value of the position of optimum individual in elite population as neural network.
To sum up it is excellent to improve whale by the way that elite retention strategy and crossover operation is added to whale optimization algorithm for the present embodiment
Change the ability of searching optimum of algorithm, and maintain optimization algorithm to be restrained, available optimal a body position pair
Neural network optimizes, and improves the parameter training ability of neural network.
Based on a upper embodiment, the present embodiment can also provide a kind of iterative processing of more specifical whale optimization algorithm
Process.
The treatment process may include：
Whale group update processing：
3 steps are broadly divided into, i.e., encirclement prey, hunting mechanism, search prey, mathematical model are as follows：
(1) prey is surrounded
Them are surrounded when humpback discovery prey, the position of optimal solution is preparatory unknown, WOA algorithm in search space
First assume that current optimal solution is exactly target prey.After target prey (the optimal particle of fitness) determines, other particles will
Position is updated along the direction of optimal particle, the mathematical model of the process is：
1)
2)
3) A=2ar1-a
4) C=2r2
Wherein, A and C indicates coefficient vector；D indicate whale individual X location updating front distance random individual position multiplied by
The length of coefficient C；T indicates current iteration number；For the position where the optimal individual of fitness in the t times iteration population
It sets；XtFor the location of whale individual in the t times iteration；A is linearly reduced to 0 from 2 with the increase of the number of iterations；r1,r2?
For the random number between [0,1].
(2) mechanism is hunted
Humpback can take contraction to surround strategy when hunting, by formula 3) the value range of A can be corresponding with the reduction of a
Subtract, the arbitrary value between A ∈ [- a, a].When the arbitrary value of A ∈ [- 1,1], passing through formula 2) the new position that can obtain humpback can be with
In its home position and current population optimal locationBetween any position.
Humpback can also take spiral parade strategy when hunting, the mathematical model for simulating the screw of humpback is：
5)
6)
Wherein, DΔIndicate the distance between whale and prey, b is constant, and for defining spiral shape, l indicates (- 1,1)
In random number.
(3) prey is searched for
When | A | when > 1, humpback can take stochastic search pattern according to mutual position, and mathematical model is as follows：
7) D=| CXrand-Xt|
8)Xt+1=Xrand-A·D
Wherein, D indicates length of the whale individual X in location updating front distance random individual position multiplied by coefficient C, XrandIt is
The location of randomly selected whale；WOA algorithm can promote other whales far from X at this timerandTo reinforce whale optimization algorithm
Ability of searching optimum.
Crossover operation may include：
It, be first the different bounds of all whale individuals, each dimension variable of different dimensions before executing crossed longitudinally operation
Normalized is executed, normalized process is as follows：
9)
Wherein, LmaxFor whale XiDimension；xi,lFor whale XiL tie up variable；Respectively whale group
The upper and lower bound of l dimension；Di,lFor xi,lCorresponding scalar after normalization.
It can be by adjusting crossover probability P when executing crossed longitudinally operationvControl the quantity of cross dimension in whale group, in order to
Part dimension is assisted to reduce the probability for destroying and normally tieing up to the greatest extent when getting rid of local optimum, each crossover operation only generates 1 filial generation, should
The model of process is as follows：
10)
11)
In formula：Ni,l1、Ni,l2The l of respectively i-th article whale1Peacekeeping l2Dimension；Random number of the r between 0-1；Si,lFor
Whale SiL ties up variable after executing renormalization operation,Indicate the intermediate variable of crossover operation process.
Processing is updated based on the whale group and the whale optimization algorithm of crossover operation may include：
Step 1, it is assumed that the parameter (weight and threshold values) of the neural network of forecasted electricity market price to share N number of, then enable every whale
For N-dimensional column vector, the position of solution is the food position of whale.Set whale group size Q, maximum number of iterations tmax, random initial
Change whale population；
Step 2, pass through formula 12) calculate whale individual fitness；
12)
Wherein, pt、Respectively practical electricity price and output electricity price, N is training sample；
Step 3, processing is updated according to whale group and updates whale group position, generated filial generation is merged into parent mixes whale
The shoal of fish filters out Q offspring individual according to elite retention strategy；
Step 4, crossover operation is executed to by the updated whale group of step 3, generates new filial generation whale group and parent
It is merged into mixed population, and Q offspring individual is filtered out by elite retention strategy and enters next iteration；
Step 5, judge whether to reach maximum number of iterations tmax.If reaching maximum number of iterations, whale group individual is exported
Optimal position is biased as the input weight and hidden layer of neural network, otherwise, executes step 3.
Based on above embodiments, the present embodiment does one and illustrates mainly for how to carry out prediction processing, other portions
Divide and be substantially the same with above embodiments, same section can refer to above embodiments, and this will not be repeated here.
Referring to FIG. 3, Fig. 3 is the flow chart of the prediction processing of Electricity price forecasting solution provided by the embodiment of the present application.
The processing may include：
S301 carries out resolution process to original electricity price sequence and obtains multiple subsequences；
This step is intended to carry out resolution process to original electricity price sequence, obtains multiple subsequences.Namely original electricity price sequence
Column are decomposed into several more stable subsequences, which is normalized, so that subsequence is full
The performance requirement of sufficient neural network activation primitive.
Optionally, this step may include：
Rule of thumb pattern decomposition algorithm carries out resolution process to original electricity price sequence, obtains multiple subsequences.
Original electricity price sequence is decomposed using the algorithm of empirical mode decomposition namely in the optinal plan, it is acquired
Subsequence it is also just more steady, prediction processing is more conducively carried out by prediction neural network.
S302 carries out prediction processing to all subsequences using prediction neural network, obtains multiple subsequence predicted values；
On the basis of step S301, this step is intended to carry out at prediction all subsequences using prediction neural network
Reason, the subsequence after equally also being predicted accordingly for subsequence prediction, that is, subsequence predicted value.
Wherein, any one neural network that the method for the prediction processing in this step can be provided using the prior art is pre-
Survey method, this will not be repeated here.
All subsequence predicted values are overlapped to obtain Research on electricity price prediction result by S303.
On the basis of step S302, this step is intended to be overlapped to obtain Research on electricity price prediction knot by all subsequence predicted values
Fruit.
By the neural network that previous embodiment obtains neural network is trained up to parameter, to mention
To the predictive ability of original electricity price sequence in high the present embodiment, the accuracy of prediction is improved.
A kind of Research on electricity price prediction device based on whale optimization algorithm provided by the embodiments of the present application is introduced below, under
A kind of Research on electricity price prediction device based on whale optimization algorithm of text description and above-described a kind of based on whale optimization algorithm
Electricity price forecasting solution can correspond to each other reference.
Specifically, referring to FIG. 4, Fig. 4 is a kind of electricity price based on whale optimization algorithm provided by the embodiment of the present application
The structural schematic diagram of prediction meanss.
The apparatus may include：
Neural metwork training module 100, for the knot according to elite retention strategy and crossover operation to whale optimization algorithm
Fruit is iterated optimization processing, obtains the weight and threshold value of neural network；According to weight and threshold value to the initial nerve net of building
Network carries out parameter adjustment, obtains prediction neural network；
Research on electricity price prediction module 200, for being pre-processed to original electricity price sequence, using prediction neural network to pretreatment
As a result prediction processing is carried out, Research on electricity price prediction result is obtained.
The neural metwork training module 100, including iterative optimization unit and parameter adjustment unit；, wherein iteration optimization list
Member may include：Including：
Initialization process unit obtains initialization population for carrying out random initializtion processing to whale optimization algorithm；
Whale group location updating unit, for being carried out at whale group update according to whale optimization algorithm to initialization population
Obtained progeny population and initialization population are merged into mixed population by reason, are carried out according to elite retention strategy to mixed population
Screening Treatment obtains elite population；
Population cross processing unit, for executing crossover operation to elite population, by obtained intersection progeny population and essence
English population merges into cross-mixing population, carries out Screening Treatment to cross-mixing population according to elite retention strategy, is intersected
Elite population；
Iteration result acquiring unit, for that will intersect elite population as initialization population and execute corresponding processing, directly
To execute number reach maximum number of iterations, using intersect elite population in optimum individual position as the weight of neural network with
Threshold value.
The Research on electricity price prediction module 200 may include：
Pretreatment unit obtains multiple subsequences for carrying out resolution process to original electricity price sequence；
Predicting unit obtains multiple subsequences for carrying out prediction processing to all subsequences using prediction neural network
Predicted value；
As a result acquiring unit, for being overlapped to obtain Research on electricity price prediction result by all subsequence predicted values.
Wherein, pretreatment unit can be used for rule of thumb pattern decomposition algorithm and carry out at decomposition to original electricity price sequence
Reason, obtains multiple subsequences.
The embodiment of the present application also provides a kind of server, including：
Memory, for storing computer program；
Processor, when for executing computer program the step of the realization such as Electricity price forecasting solution of above embodiments.
The embodiment of the present application also provides a kind of computer readable storage medium, and meter is stored on computer readable storage medium
Calculation machine program, when computer program is executed by processor the step of the realization such as Electricity price forecasting solution of above embodiments.
The computer readable storage medium may include：USB flash disk, mobile hard disk, read-only memory (Read-Only
Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. is various to deposit
Store up the medium of program code.
Each embodiment is described in a progressive manner in specification, the highlights of each of the examples are with other realities
The difference of example is applied, the same or similar parts in each embodiment may refer to each other.For device disclosed in embodiment
Speech, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is referring to method part illustration
?.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond scope of the present application.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
Above to a kind of Electricity price forecasting solution based on whale optimization algorithm provided herein, Research on electricity price prediction device,
Server and computer readable storage medium are described in detail.Principle of the specific case to the application used herein
And embodiment is expounded, the present processes that the above embodiments are only used to help understand and its core are thought
Think.It should be pointed out that for those skilled in the art, under the premise of not departing from the application principle, may be used also
With to the application, some improvement and modification can also be carried out, these improvement and modification are also fallen into the protection scope of the claim of this application.
Claims (10)
1. a kind of Electricity price forecasting solution based on whale optimization algorithm, which is characterized in that including：
Optimization processing is iterated to the result of whale optimization algorithm according to elite retention strategy and crossover operation, obtains nerve net
The weight and threshold value of network；Parameter adjustment is carried out to the initial neural network of building according to the weight and threshold value, obtains prediction mind
Through network；
Original electricity price sequence is pre-processed, prediction processing is carried out to pre-processed results using the prediction neural network, is obtained
To Research on electricity price prediction result.
2. Electricity price forecasting solution according to claim 1, which is characterized in that according to elite retention strategy and crossover operation pair
The result of whale optimization algorithm is iterated optimization processing, obtains the weight and threshold value of neural network, including：
Random initializtion processing is carried out to whale optimization algorithm, obtains initialization population；
Whale group update processing is carried out to the initialization population according to whale optimization algorithm, by obtained progeny population and initially
Change population and merge into mixed population, Screening Treatment is carried out to the mixed population according to the elite retention strategy and obtains elite kind
Group；
Crossover operation is executed to the elite population, obtained intersection progeny population is merged into intersect with the elite population and is mixed
Population is closed, Screening Treatment is carried out to the cross-mixing population according to the elite retention strategy, obtains intersecting elite population；
The intersection elite population as the initialization population and is executed into corresponding processing, reaches maximum until executing number
The number of iterations, using the position of optimum individual in the intersection elite population as the weight of neural network and threshold value.
3. Electricity price forecasting solution according to claim 1, which is characterized in that pre-process, adopt to original electricity price sequence
Prediction processing is carried out to pre-processed results with the prediction neural network, obtains Research on electricity price prediction as a result, including：
Resolution process is carried out to the original electricity price sequence and obtains multiple subsequences；
Prediction processing is carried out to all subsequences using the prediction neural network, obtains multiple subsequence predicted values；
All subsequence predicted values are overlapped to obtain Research on electricity price prediction result.
4. Electricity price forecasting solution according to claim 3, which is characterized in that carried out at decomposition to the original electricity price sequence
Reason obtains multiple subsequences, including：
Rule of thumb pattern decomposition algorithm carries out resolution process to the original electricity price sequence, obtains the multiple subsequence.
5. a kind of Research on electricity price prediction device based on whale optimization algorithm, which is characterized in that including：
Neural metwork training module, for being changed according to elite retention strategy and crossover operation to the result of whale optimization algorithm
For optimization processing, the weight and threshold value of neural network are obtained；According to the weight and threshold value to the initial neural network of building into
The adjustment of row parameter, obtains prediction neural network；
Research on electricity price prediction module ties pretreatment using the prediction neural network for pre-processing to original electricity price sequence
Fruit carries out prediction processing, obtains Research on electricity price prediction result.
6. Research on electricity price prediction device according to claim 5, which is characterized in that the neural metwork training module, including repeatedly
Generation optimization unit and parameter adjustment unit；
Wherein, the iterative optimization unit includes：
Initialization process subelement obtains initialization population for carrying out random initializtion processing to whale optimization algorithm；
Whale group location updating subelement, for being carried out at whale group update according to whale optimization algorithm to the initialization population
Obtained progeny population and initialization population are merged into mixed population, according to the elite retention strategy to the mixing by reason
Population carries out Screening Treatment and obtains elite population；
Population cross processing subelement, for the elite population execute crossover operation, by obtained intersection progeny population with
The elite population merges into cross-mixing population, is screened according to the elite retention strategy to the cross-mixing population
Processing obtains intersecting elite population；
Iteration result obtains subelement, for the intersection elite population as the initialization population and to be executed corresponding place
Reason reaches maximum number of iterations until executing number, using the position of optimum individual in the intersection elite population as nerve net
The weight and threshold value of network.
7. Research on electricity price prediction device according to claim 5, which is characterized in that the Research on electricity price prediction module, including：
Pretreatment unit obtains multiple subsequences for carrying out resolution process to the original electricity price sequence；
Predicting unit obtains multiple sons for carrying out prediction processing to all subsequences using the prediction neural network
Sequence prediction value；
As a result acquiring unit, for being overlapped to obtain Research on electricity price prediction result by all subsequence predicted values.
8. Research on electricity price prediction device according to claim 7, which is characterized in that the pretreatment unit is specifically used for according to warp
It tests pattern decomposition algorithm and resolution process is carried out to the original electricity price sequence, obtain the multiple subsequence.
9. a kind of server, which is characterized in that including：
Memory, for storing computer program；
Processor realizes such as Claims 1-4 described in any item Electricity price forecasting solutions when for executing the computer program
The step of.
10. a kind of computer readable storage medium, which is characterized in that be stored with computer on the computer readable storage medium
Program is realized when the computer program is executed by processor such as the described in any item Electricity price forecasting solutions of Claims 1-4
Step.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810877725.7A CN108830431A (en) | 2018-08-03 | 2018-08-03 | A kind of Electricity price forecasting solution and relevant apparatus based on whale optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810877725.7A CN108830431A (en) | 2018-08-03 | 2018-08-03 | A kind of Electricity price forecasting solution and relevant apparatus based on whale optimization algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108830431A true CN108830431A (en) | 2018-11-16 |
Family
ID=64153487
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810877725.7A Pending CN108830431A (en) | 2018-08-03 | 2018-08-03 | A kind of Electricity price forecasting solution and relevant apparatus based on whale optimization algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108830431A (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109492828A (en) * | 2018-12-12 | 2019-03-19 | 河南科技学院 | A kind of Distribution path optimization method for considering customer grade and distribution time and requiring |
| CN109765893A (en) * | 2019-01-17 | 2019-05-17 | 重庆邮电大学 | Method for planning path for mobile robot based on whale optimization algorithm |
| CN109886589A (en) * | 2019-02-28 | 2019-06-14 | 长安大学 | A method of low-carbon Job-Shop is solved based on whale optimization algorithm is improved |
| CN109940610A (en) * | 2019-02-22 | 2019-06-28 | 浙江工业大学 | A kind of joint of robot control moment prediction technique based on WOA-GA hybrid optimization algorithm |
| CN110110930A (en) * | 2019-05-08 | 2019-08-09 | 西南交通大学 | A kind of Recognition with Recurrent Neural Network Short-Term Load Forecasting Method improving whale algorithm |
| CN110322050A (en) * | 2019-06-04 | 2019-10-11 | 西安邮电大学 | A kind of wind energy resources compensation data method |
| CN112581262A (en) * | 2020-12-23 | 2021-03-30 | 百维金科(上海)信息科技有限公司 | Whale algorithm-based fraud detection method for optimizing LVQ neural network |
| CN112700060A (en) * | 2021-01-08 | 2021-04-23 | 佳源科技股份有限公司 | Station terminal load prediction method and prediction device |
| CN113656266A (en) * | 2021-07-26 | 2021-11-16 | 广东奥博信息产业股份有限公司 | Performance prediction method and system for government-enterprise service system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106485365A (en) * | 2016-10-25 | 2017-03-08 | 广东工业大学 | A kind of Load Prediction In Power Systems method and device |
| CN107016436A (en) * | 2017-03-31 | 2017-08-04 | 浙江大学 | A kind of mixing whale algorithm of bionical policy optimization |
| CN107480829A (en) * | 2017-08-25 | 2017-12-15 | 广东工业大学 | A kind of Short-term electricity price forecasting method, apparatus and system |
| CN108205698A (en) * | 2017-06-15 | 2018-06-26 | 广东工业大学 | A kind of cloud resource load predicting method based on the double string whale optimization algorithms of just remaining chaos |
-
2018
- 2018-08-03 CN CN201810877725.7A patent/CN108830431A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106485365A (en) * | 2016-10-25 | 2017-03-08 | 广东工业大学 | A kind of Load Prediction In Power Systems method and device |
| CN107016436A (en) * | 2017-03-31 | 2017-08-04 | 浙江大学 | A kind of mixing whale algorithm of bionical policy optimization |
| CN108205698A (en) * | 2017-06-15 | 2018-06-26 | 广东工业大学 | A kind of cloud resource load predicting method based on the double string whale optimization algorithms of just remaining chaos |
| CN107480829A (en) * | 2017-08-25 | 2017-12-15 | 广东工业大学 | A kind of Short-term electricity price forecasting method, apparatus and system |
Non-Patent Citations (2)
| Title |
|---|
| 杜沛: "基于多目标鲸鱼优化算法和Elman神经网络的短期风速预测模型的研究与应用", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
| 许瑜飞等: "改进鲸鱼优化算法及其在渣油加氢参数优化的应用", 《化工学报》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109492828A (en) * | 2018-12-12 | 2019-03-19 | 河南科技学院 | A kind of Distribution path optimization method for considering customer grade and distribution time and requiring |
| CN109492828B (en) * | 2018-12-12 | 2021-08-13 | 河南科技学院 | Distribution path optimization method considering customer grade and distribution time requirements |
| CN109765893A (en) * | 2019-01-17 | 2019-05-17 | 重庆邮电大学 | Method for planning path for mobile robot based on whale optimization algorithm |
| CN109940610A (en) * | 2019-02-22 | 2019-06-28 | 浙江工业大学 | A kind of joint of robot control moment prediction technique based on WOA-GA hybrid optimization algorithm |
| CN109940610B (en) * | 2019-02-22 | 2020-10-30 | 浙江工业大学 | Robot joint control moment prediction method based on WOA-GA (weighted average-genetic algorithm) hybrid optimization algorithm |
| CN109886589A (en) * | 2019-02-28 | 2019-06-14 | 长安大学 | A method of low-carbon Job-Shop is solved based on whale optimization algorithm is improved |
| CN109886589B (en) * | 2019-02-28 | 2024-01-05 | 长安大学 | Method for solving low-carbon workshop scheduling based on improved whale optimization algorithm |
| CN110110930A (en) * | 2019-05-08 | 2019-08-09 | 西南交通大学 | A kind of Recognition with Recurrent Neural Network Short-Term Load Forecasting Method improving whale algorithm |
| CN110110930B (en) * | 2019-05-08 | 2022-03-25 | 西南交通大学 | Recurrent neural network short-term power load prediction method for improving whale algorithm |
| CN110322050A (en) * | 2019-06-04 | 2019-10-11 | 西安邮电大学 | A kind of wind energy resources compensation data method |
| CN110322050B (en) * | 2019-06-04 | 2023-04-07 | 西安邮电大学 | Wind energy resource data compensation method |
| CN112581262A (en) * | 2020-12-23 | 2021-03-30 | 百维金科(上海)信息科技有限公司 | Whale algorithm-based fraud detection method for optimizing LVQ neural network |
| CN112700060A (en) * | 2021-01-08 | 2021-04-23 | 佳源科技股份有限公司 | Station terminal load prediction method and prediction device |
| CN113656266A (en) * | 2021-07-26 | 2021-11-16 | 广东奥博信息产业股份有限公司 | Performance prediction method and system for government-enterprise service system |
| CN113656266B (en) * | 2021-07-26 | 2023-10-27 | 广东奥博信息产业股份有限公司 | Performance prediction method and system for government enterprise service system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108830431A (en) | A kind of Electricity price forecasting solution and relevant apparatus based on whale optimization algorithm | |
| Cho et al. | Basic enhancement strategies when using Bayesian optimization for hyperparameter tuning of deep neural networks | |
| CN108345937B (en) | Circulation is merged with library | |
| CN111105029B (en) | Neural network generation method, generation device and electronic equipment | |
| Zhao et al. | Toward SLA-constrained service composition: An approach based on a fuzzy linguistic preference model and an evolutionary algorithm | |
| CN110462638A (en) | Training neural network is sharpened using posteriority | |
| CN109670101A (en) | Crawler dispatching method, device, electronic equipment and storage medium | |
| CN108846695A (en) | The prediction technique and device of terminal replacement cycle | |
| WO2020188341A1 (en) | Adaptive clinical trials | |
| CN108831561A (en) | Generation method, device and the computer readable storage medium of influenza prediction model | |
| CN108647064A (en) | The method and device of courses of action navigation | |
| CN116992151A (en) | Online course recommendation method based on double-tower graph convolution neural network | |
| CN108846687A (en) | Client segmentation method, apparatus and storage medium | |
| Alnowibet et al. | An efficient algorithm for data parallelism based on stochastic optimization | |
| Xu et al. | Two sided disassembly line balancing problem with rest time of works: A constraint programming model and an improved NSGA II algorithm | |
| Sadeghiram et al. | Multi-objective distributed Web service composition—a link-dominance driven evolutionary approach | |
| CN111898766B (en) | Ether house fuel limitation prediction method and device based on automatic machine learning | |
| Zhang et al. | ReLP: reinforcement learning pruning method based on prior knowledge | |
| CN117539648A (en) | Service quality management method and device for electronic government cloud platform | |
| WO2021139255A1 (en) | Model based method and apparatus for predicting data change frequency, and computer device | |
| Li et al. | Fast randomized algorithm with restart strategy for minimal test cost feature selection | |
| Ingber | Ideas by statistical mechanics (ISM) | |
| CN111027709B (en) | Information recommendation method and device, server and storage medium | |
| Crowley et al. | Policy Gradient Planning for Environmental Decision Making with Existing Simulators. | |
| CN113420877A (en) | Financial data decision method and device, electronic equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181116 |
|
| RJ01 | Rejection of invention patent application after publication |