CN116499023A - Intelligent control method and system for geothermal coupling solar heating station - Google Patents
Intelligent control method and system for geothermal coupling solar heating station Download PDFInfo
- Publication number
- CN116499023A CN116499023A CN202310451141.4A CN202310451141A CN116499023A CN 116499023 A CN116499023 A CN 116499023A CN 202310451141 A CN202310451141 A CN 202310451141A CN 116499023 A CN116499023 A CN 116499023A
- Authority
- CN
- China
- Prior art keywords
- data
- operation parameters
- prediction model
- algorithm
- fault
- 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
- 238000010438 heat treatment Methods 0.000 title claims abstract description 67
- 230000008878 coupling Effects 0.000 title claims abstract description 61
- 238000010168 coupling process Methods 0.000 title claims abstract description 61
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 61
- 238000000034 method Methods 0.000 title claims abstract description 47
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 104
- 238000003745 diagnosis Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 65
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 48
- 238000012795 verification Methods 0.000 claims description 32
- 238000010248 power generation Methods 0.000 claims description 28
- 238000007637 random forest analysis Methods 0.000 claims description 26
- 230000006870 function Effects 0.000 claims description 25
- 238000001816 cooling Methods 0.000 claims description 16
- 230000002159 abnormal effect Effects 0.000 claims description 14
- 238000012544 monitoring process Methods 0.000 claims description 11
- 238000004146 energy storage Methods 0.000 claims description 10
- 238000010200 validation analysis Methods 0.000 claims description 9
- 230000002787 reinforcement Effects 0.000 claims description 8
- 238000002790 cross-validation Methods 0.000 claims description 7
- 238000012423 maintenance Methods 0.000 claims description 7
- 238000010219 correlation analysis Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000003044 adaptive effect Effects 0.000 claims description 4
- 238000003066 decision tree Methods 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000001133 acceleration Effects 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims description 2
- 238000010801 machine learning Methods 0.000 abstract description 2
- 239000002689 soil Substances 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000001105 regulatory effect Effects 0.000 description 4
- 230000033228 biological regulation Effects 0.000 description 3
- 239000002131 composite material Substances 0.000 description 2
- 230000001276 controlling effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005057 refrigeration Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 125000004122 cyclic group Chemical group 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005553 drilling Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000008929 regeneration Effects 0.000 description 1
- 238000011069 regeneration method Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
- 238000010977 unit operation Methods 0.000 description 1
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24D—DOMESTIC- OR SPACE-HEATING SYSTEMS, e.g. CENTRAL HEATING SYSTEMS; DOMESTIC HOT-WATER SUPPLY SYSTEMS; ELEMENTS OR COMPONENTS THEREFOR
- F24D19/00—Details
- F24D19/10—Arrangement or mounting of control or safety devices
- F24D19/1006—Arrangement or mounting of control or safety devices for water heating systems
- F24D19/1009—Arrangement or mounting of control or safety devices for water heating systems for central heating
- F24D19/1039—Arrangement or mounting of control or safety devices for water heating systems for central heating the system uses a heat pump
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24D—DOMESTIC- OR SPACE-HEATING SYSTEMS, e.g. CENTRAL HEATING SYSTEMS; DOMESTIC HOT-WATER SUPPLY SYSTEMS; ELEMENTS OR COMPONENTS THEREFOR
- F24D15/00—Other domestic- or space-heating systems
- F24D15/02—Other domestic- or space-heating systems consisting of self-contained heating units, e.g. storage heaters
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24D—DOMESTIC- OR SPACE-HEATING SYSTEMS, e.g. CENTRAL HEATING SYSTEMS; DOMESTIC HOT-WATER SUPPLY SYSTEMS; ELEMENTS OR COMPONENTS THEREFOR
- F24D15/00—Other domestic- or space-heating systems
- F24D15/04—Other domestic- or space-heating systems using heat pumps
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24D—DOMESTIC- OR SPACE-HEATING SYSTEMS, e.g. CENTRAL HEATING SYSTEMS; DOMESTIC HOT-WATER SUPPLY SYSTEMS; ELEMENTS OR COMPONENTS THEREFOR
- F24D19/00—Details
- F24D19/10—Arrangement or mounting of control or safety devices
- F24D19/1006—Arrangement or mounting of control or safety devices for water heating systems
- F24D19/1009—Arrangement or mounting of control or safety devices for water heating systems for central heating
- F24D19/1042—Arrangement or mounting of control or safety devices for water heating systems for central heating the system uses solar energy
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
- G06Q10/06375—Prediction of business process outcome or impact based on a proposed change
-
- 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
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- General Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Theoretical Computer Science (AREA)
- Combustion & Propulsion (AREA)
- Chemical & Material Sciences (AREA)
- Thermal Sciences (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Sustainable Energy (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Sustainable Development (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses an intelligent control method and system for a geothermal coupling solar heating station, and belongs to the technical field of operation management of a soil source heat pump. After fault diagnosis is carried out according to current operation parameters, an integrated algorithm prediction model is constructed according to historical operation parameters through a machine learning algorithm, the load of a next time step system is output according to current meteorological conditions, decisions are made according to the prediction results and the current operation parameters and are transmitted to a load adjusting module, the adjustment decisions are transmitted to a DCS system after being checked by the load adjusting module, the heat supply efficiency of the whole system is improved in a mode of changing operation conditions in advance, meanwhile, the influence of solar grid connection on the stability of a power grid is reduced, and the stability of the whole system is improved.
Description
Technical Field
The invention belongs to the technical field of operation management of a ground source heat pump, and particularly relates to an intelligent control method and system of a geothermal coupling solar heating station.
Background
Efficient and clean utilization of energy has been a goal pursued in various countries. Clean energy which can be continuously developed and utilized is rapidly developed in recent years, meanwhile, fossil energy on the earth is gradually consumed, and new energy development and utilization are imperative.
The solar energy and the geothermal energy have the characteristics of continuous and stable energy supply, high-efficiency cyclic utilization and regeneration, and have wide application in industries such as heating, power generation and the like. However, the renewable energy industry has the problems of extensive development and low utilization efficiency. With the comprehensive promotion of new technologies such as big data and artificial intelligence in industry, the change of a simple rough utilization mode of renewable energy sources is urgent in the past, and the demands for the refinement and the intellectualization of the energy sources are urgent.
In order to reduce carbon emission and improve energy utilization efficiency, renewable energy sources are used for generating power in a grid-connected mode while supplying heat, but the safety and the efficiency of a heat supply system are greatly affected by the intermittent performance and the change of the heat load of a user side, so that the heat supply unit must use an energy storage technology to improve the load matching capability. The problems of heat supply load change, load mismatch and the like can seriously affect the high efficiency and the safety of unit operation, so that the improvement of the response capability of a heat supply system to the load change is urgent.
Disclosure of Invention
In order to solve the defects in the prior art, the invention aims to provide an intelligent control method and system for a geothermal coupling solar heating station, which can predict the operation parameters of an energy system in advance and make corresponding adjustment, thereby remarkably improving the efficiency of the whole system.
The invention is realized by the following technical scheme:
an intelligent control method for a geothermal coupling solar heating station comprises the following steps:
s1: monitoring the operation parameters of the geothermal coupling solar heating station according to a preset monitoring period, and performing fault diagnosis on the obtained operation parameters; if the operation parameters are abnormal, comparing the operation parameters with the historical data, and determining a fault solution according to the comparison result; if the operation parameters are normal, turning to S2;
s2: taking the historical normal operation parameters of the geothermal coupling solar heating station and the corresponding historical meteorological conditions as data sets, dividing the data sets into training sets and verification sets, and constructing an integrated algorithm prediction model; training an integrated algorithm prediction model by using an integrated prediction algorithm based on the combination of a nearest neighbor node algorithm, a random forest algorithm and a reinforcement learning algorithm in a training set, optimizing the integrated algorithm prediction model by using a verification set, and predicting the requirements of a user side and the power generation capacity condition of a photovoltaic system;
s3: according to the normal operation parameters obtained in the step S1 and the prediction results of the step S2, the operation conditions of all parts of the geothermal coupling solar heating station are decided by calculating the checking flow, the cold and hot loads and the generated energy;
S4: calculating the decided operation parameters according to the decision result obtained in the step S3, comparing the decided operation parameters with fault data, and transmitting the adjusting signals to a DCS (distributed control system) of the geothermal coupling solar heat supply station after confirming that the power supply system and the heat supply system of the geothermal coupling solar heat supply station can normally operate; if the power supply system and the heating system of the geothermal coupling solar heating station cannot normally operate after adjustment, the adjustment signal is ignored to maintain the original working condition, and error adjustment data are uploaded.
Preferably, in S1, the operation parameters include a selected time point, an instantaneous temperature at the corresponding time point, a highest air temperature on the day, a lowest air temperature on the day, a solar time on the day, an air humidity on the day, an average wind speed on the day, and a date;
normalizing the operation parameters before fault diagnosis:
wherein x is i,j Representing a j-th parameter in the i-th set of data; x's' i,j Represents x i,j Normalizing the obtained parameters; min (x) ,j ) Represents the minimum value, max (x ,j ) Represent the firstThe maximum of j parameters in each group.
Preferably, in S2, the training method of the prediction model includes:
s2.1: according to the historical normal operation parameters of the geothermal coupling solar heating station, importing corresponding historical meteorological conditions, and carrying out normalization processing on the historical normal operation parameters and the historical meteorological conditions;
S2.2: analyzing main factors influencing heat load, cold load and generating capacity of a photovoltaic system by using a correlation analysis method, removing low-influence data, and taking the rest data as a data set;
s2.3: dividing the data set into a training set and a verification set by using a cross verification method;
s2.4: constructing an integrated algorithm prediction model according to the training set and the verification set, respectively training a nearest neighbor algorithm prediction model, a random forest algorithm prediction model and a self-adaptive lifting algorithm prediction model, and weighting and outputting a prediction result to a final prediction result;
s2.5: the integrated algorithm prediction model updates corresponding weights according to the prediction results of the nearest neighbor algorithm prediction model, the random forest algorithm prediction model and the self-adaptive lifting algorithm prediction model and the deviation back propagation of the actual operation parameters of the geothermal coupling solar heating station; retraining the integrated algorithm prediction model constructed in the step S2.4 when the prediction result and the operation parameter of the integrated algorithm prediction model continuously deviate for a plurality of times within a certain time;
s2.6: and repeating the steps S2.3-S2.5 to respectively obtain a user side heat demand prediction model, a user side cold demand prediction model and a photovoltaic system power generation amount prediction model.
Further preferably, S2.2 specifically includes:
s2.2.1: according to a preset time step, respectively selecting a user thermal load, a user cold load and the power generation amount of the photovoltaic array at the starting time point of each time step as a master sequence by using a gray correlation method; the subsequence selects a time point, an instantaneous temperature at a corresponding time point, a highest air temperature on the day, a lowest air temperature on the day, a sunshine time on the day, an air humidity on the day, an average wind speed on the day and a date;
s2.2.2: according to the temperature and sunlight time which are inversely related to the heat load, multiplying the heat load by a reciprocal operator in the heat load correlation analysis:
X i 'D 1 ＝[x i ' ,1 d 1 ,x i ' ,2 d 2 ,...,x i ' ,j d j ]
wherein X is i ' represents the ith subsequence; x is x i ' ,j Representing a j-th parameter in the i-th set of data; d (D) 1 Is a reciprocal operator;
s2.2.3: gray correlation degrees of all influence factors on the user side heat requirement, the user side cold requirement and the photovoltaic array power generation amount are respectively obtained through a gray correlation method, and low correlation factors except for time points are removed from the influence factors; and the processed data is used as a data set R, expressed as:
further preferably, S2.3 specifically includes:
s2.3.1: dividing the data set R obtained by processing in the step S2.2 by using a cross validation method, selecting 5-fold cross validation, dividing the data set R into 5 parts, wherein 4 parts are used as training sets and 1 part is used as validation set;
S2.3.2: step S2.3.1 is repeated 5 times, and different training sets are selected each time, so that 20 training sets and 5 verification sets are obtained.
Further preferably, S2.4 specifically includes:
s2.4.1: establishing a nearest neighbor algorithm prediction model, setting a super parameter k, selecting Manhattan distance as distance measurement, selecting mean square error and average absolute error as loss functions, selecting Gaussian functions as weights based on distance, and obtaining the nearest neighbor algorithm prediction model based on a training set and a verification set obtained in the step S2.3, taking a minimized loss function as a target, and using KD tree acceleration training; expressed as:
wherein y is i Is the true value of the i-th parameter,the predicted value of the ith parameter, n is the number of dimensions;
wherein a, b, c are parameters, x is the distance from a point in space to a true value;
s2.4.2: establishing a random forest regression prediction model, and initially selecting N tree Setting the maximum depth of each feature tree to be 50, and initially selecting the feature value of each feature tree to be m try The method comprises the steps of carrying out a first treatment on the surface of the Selecting the data set R obtained in the step S2.2, and using a bagging method to carry out put-back sampling to generate N tree Training subsets, wherein the data outside the bags which are not extracted are used as verification sets; dividing the feature tree by CART method, and randomly selecting m from j influencing factors try A influencing factor (m) try J) as the splitting characteristic value of the current node of the characteristic tree, selecting a mean square error and average absolute error minimization criterion as the splitting standard of the characteristic tree, and predicting the result as the average value output by each tree; iteratively calculating the quantity of the optimized feature trees and the feature values of the decision trees with the minimum error as a target until the error is smaller than a set threshold value to obtain a random forest prediction model;
s2.4.3: establishing a self-adaptive lifting algorithm prediction model, selecting a training set obtained in the step S2.3, selecting n groups of training samples from the training set, and endowing the training samples with initial weights of w 1i =1/n, i=1, 2,3, …, n, with W (1) = (W) 11 ,w 12 ,w 13 ,…,w 1n ) Representing the initial weight of the sample, D representing the number of learners; performing the iteration d=1, 2,3, …, D, training the D-th weak learner H d (x) When using weak learner H d (x) The prediction validation set outputs regression error epsilon d Calculating the maximum error E of samples on the training set d And relative error epsilon di Calculating the weight alpha of the weak learner in the final learner d According to the weight alpha d Updating the weight w (d+1) of the sample; training D wheels to obtain D groups of weak learners H d (x) Combining the weak learners according to the weak learner weights to obtain an enhanced learner h (x), and optimizing the weak learner number D by using the verification set data; expressed as:
E d ＝max(|y i -H d (x i )|)
W(d+1)＝(w d|1,1 ,w d|1,2 ,w d|1,3 ,…,w d|1,n )
Wherein y is i Is the true value of the i-th parameter.
Further preferably, S2.5 specifically includes:
s2.5.1: optimizing the integrated algorithm prediction model; inputting verification set data obtained in the step S2.3, respectively obtaining output by using a trained nearest neighbor algorithm prediction model, a random forest regression prediction model and an adaptive lifting algorithm prediction model, taking the mean square error of an integrated algorithm model prediction result and a true value as a loss function, and optimizing the weight proportion of each part by using historical data until the loss function J is reached θ Less than a set threshold:
D＝w 1 d 1 +w 2 d 2 +w 3 d 3
d is an output result of the integrated algorithm prediction model; d, d i ,w i The method comprises the steps of predicting a result and a weight by a nearest neighbor algorithm, a random forest algorithm and a self-adaptive lifting algorithm respectively; w (w) i ' is the updated weight; d' is a true value; k is a parameter for controlling the weight update;
s2.5.2: and uploading the prediction result of the integrated algorithm prediction model to a cloud database every time, and when the prediction result error or the fault decision in a period of time exceeds a threshold value, sending a signal by the cloud database to remind maintenance personnel to update the learning prediction module, and repeating the steps S2.1-S2.5.1 to retrain the prediction model.
The invention discloses an intelligent control system of a geothermal coupling solar heating station, which comprises the following components:
The operation checking module is used for monitoring the operation parameters of the geothermal coupling solar heating station according to a preset monitoring period and performing fault diagnosis on the obtained operation parameters; if the operation parameters are abnormal, uploading the operation parameters to a cloud database; if the operation parameters are normal, transmitting the operation parameters to a real-time decision module;
the learning prediction module reads historical normal operation parameters of the geothermal coupling solar heat supply station from the cloud database, combines corresponding historical meteorological conditions to serve as a data set, divides the data set into a training set and a verification set, and constructs an integrated algorithm prediction model; training an integrated algorithm prediction model by using an integrated prediction algorithm based on the combination of a nearest neighbor node algorithm, a random forest algorithm and a reinforcement learning algorithm and optimizing the integrated algorithm prediction model by using a verification set, predicting the conditions of the demand of a user side and the power generation capacity of a photovoltaic system, and transmitting a prediction result to a real-time decision module;
the real-time decision module is used for respectively reading the normal operation parameters from the operation checking module and the prediction results of the learning prediction module, deciding the operation conditions of each part of the geothermal coupling solar heat supply station by calculating the checking flow, the cooling and heating load and the generating capacity, transmitting the decision results to the load adjusting module, and uploading the normal operation parameters from the operation checking module and the prediction results of the learning prediction module to the cloud database;
The load adjusting module is used for reading the decision result from the real-time decision module, calculating the decided operation parameters, comparing the decided operation parameters with fault data in the cloud database, and confirming that the power supply system and the heat supply system of the geothermal coupling solar heat supply station can normally operate and then transmitting an adjusting signal to the DCS system of the geothermal coupling solar heat supply station; if the power supply system and the heating system of the geothermal coupling solar heating station cannot normally operate after adjustment, ignoring the adjustment signal to maintain the original working condition, and uploading error adjustment data to a cloud database;
the cloud database is used for storing normal operation parameters, fault operation parameters, prediction result data and fault decision data and comparing the data with uploaded data.
Preferably, the monitoring period of the operation checking module, the decision period of the learning prediction module, the decision period of the real-time decision module and the adjustment period of the load adjustment module are consistent with a preset time step.
Preferably, the operation checking module monitors the operation parameters of the geothermal coupling solar heating station, including the water temperature and flow rate of the inlet and outlet of the user end, the water temperature and flow rate of the inlet and outlet of the deep buried pipe, the water temperature and flow rate of the shallow buried pipe, the water temperature and flow rate of the inlet and outlet of the cooling tower, the temperature and water storage condition of the underground water storage tank, the power consumption of the geothermal energy system, the power generation power of the photovoltaic array and the electric quantity storage condition of the energy storage battery, and records the current operation parameters in the forms of date, time period, device and working condition; after the operation checking module presets an abnormal parameter range to obtain the current operation parameter, firstly judging whether the flow, the energy and the electric quantity of a power supply system and a heat supply system of the geothermal coupling solar heat supply station are conservation or not based on the current operation mode, and carrying out fault judgment by comparing the preset abnormal parameter range; when possible abnormal data exist, the operation checking module transmits current fault data to the cloud database, the cloud database compares the data according to the operation parameters of three adjacent time steps and the operation parameters of the historical fault database, determines a fault position, actively alarms, generates a fault guiding scheme, and uploads the fault data and the fault position to the cloud database;
The cloud database comprises a normal operation parameter database, a fault operation parameter database, a prediction result database and a fault decision database; the cloud database has a data comparison function, and specifically comprises the steps of comparing possible fault data transmitted by the operation checking module with historical fault data, comparing the decided operation parameters calculated by the load adjusting module with the fault operation database, and comparing a prediction result with corresponding real data; when the prediction result error or the fault decision exceeds a threshold value within a period of time, the cloud database sends out a reminding signal to prompt maintenance personnel to update the learning prediction module.
Compared with the prior art, the invention has the following beneficial technical effects:
according to the geothermal coupling solar heat supply station intelligent control method disclosed by the invention, an integrated algorithm prediction model is constructed based on a machine learning principle, the cold and hot requirements of a user side of the geothermal coupling solar heat supply station and the generated energy of a photovoltaic system of the geothermal coupling solar heat supply station after a period of time are predicted according to the current meteorological conditions, and the operation parameters of the heat supply station are adjusted in advance according to the prediction results. The load is predicted and matched through the integrated algorithm prediction model, so that the utilization rate of energy sources of the geothermal coupling solar heating station can be improved, and the safety of the heating station is ensured. The integrated prediction algorithm consists of three parts, namely a nearest neighbor node algorithm, a random forest algorithm and an reinforcement learning algorithm, and the final prediction result is output in a weighted average mode after the algorithm of each part is output. Compared with the single algorithm, the stability and the accuracy of the prediction model can be improved by comparing errors of all parts of the three algorithms and updating weights. The invention monitors the operation parameters of the geothermal coupling solar heating station in real time through the DCS system, actively alarms and judges the fault position when abnormal operation parameters occur, and provides a guiding scheme. The invention also records the historical operation parameters of the thermally coupled solar heating station through the cloud database, and provides data for training a prediction model and troubleshooting.
The geothermal coupling solar heat supply station intelligent control system disclosed by the invention is simple to construct, high in automation degree and complete in functions.
Drawings
FIG. 1 is a schematic block diagram of an intelligent control method of a geothermal coupling solar heating station of the present invention;
FIG. 2 is a schematic diagram of the intelligent control system of the geothermal coupling solar heating station of the present invention;
FIG. 3 is a schematic diagram of an integrated algorithm predictive model of the present invention;
fig. 4 is a schematic structural diagram of a geothermal coupling solar heating station according to an embodiment of the present invention.
In the figure, 1 is a first water pump; 2 is a second water pump; 3 is a third water pump; 4 is a fourth water pump; 5 is a deep buried pipe; 6 is an underground water storage tank; 7 is a shallow buried pipe group; 8 is a closed cooling tower; 9 is a heating device; 10 is a photovoltaic array; 11 is a photovoltaic power generation device; 12 is an energy storage battery; 13 is a first solenoid valve; 14 is a second solenoid valve; 15 is a third solenoid valve; and 16 is a fourth solenoid valve.
Detailed Description
The invention will now be described in further detail with reference to the drawings and to specific embodiments, which are intended to be illustrative rather than limiting.
Referring to fig. 4, a typical geothermal coupling solar heating station is shown, the system comprises a photovoltaic array 10, a photovoltaic power generation device 11, an underground water storage tank 6, a heating device 9, an energy storage battery 12 and a composite buried pipe system, wherein the composite buried pipe system comprises a shallow buried pipe group 7, a deep buried pipe 5 and a closed cooling tower 8 which are connected in parallel; the whole system is also provided with a first water pump 1, a second water pump 2, a third water pump 3, a fourth water pump 4, a first electromagnetic valve 13, a second electromagnetic valve 14, a third electromagnetic valve 15 and a fourth electromagnetic valve 16, and the working condition during operation is transmitted to the DCS system. Wherein the shallow buried pipe group 7 is a buried pipe with the depth of 100-1000 m, and the deep buried pipe 5 is a buried pipe arranged at 1200-3000 m. The shallow buried pipes are serially arranged according to a zigzag shape, and the drilling interval of the shallow buried pipes is 5-10m; the deep buried pipe 5 adopts a double pipe heat exchanger, and is connected with a condenser and an evaporator to form a heat exchange unit, so that heat is taken without water.
Referring to fig. 2, a schematic diagram is formed for an intelligent control system of a geothermal coupling solar heating station of the present invention, and the intelligent control system includes an operation checking module, a learning prediction module, a real-time decision module, a load adjusting module and a cloud database.
As shown in fig. 1, the main functions of each module and the working principle between the modules are as follows:
the operation checking module is used for monitoring the operation of the heating station, performing fault diagnosis on the current operation parameters by taking a preset time step as a period, and transmitting the normal operation parameters to the real-time decision module; when possible abnormal data exist, the module can transmit current fault data to the cloud database, the cloud database compares historical data to determine fault conditions, the fault conditions are actively alarmed, a fault guiding scheme is given, operation maintainers are helped to quickly eliminate faults, and normal operation of the unit is guaranteed.
And the learning prediction module is used for reading the historical normal operation parameters of the heating station from the cloud and taking the historical normal operation parameters and the imported corresponding historical meteorological conditions as a data set, wherein the data set is divided into a training set and a verification set. The module trains a predictive model based on an integrated predictive algorithm combining nearest neighbor nodes, random forests, and reinforcement learning, and optimizes the predictive model using a validation set. The module and the operation checking module act at the same frequency, the prediction is carried out by taking a preset time step as a period, and a prediction result is transmitted to the real-time decision module.
The real-time decision module reads the normal operation parameters from the operation checking module and the prediction results of the learning prediction module as decision basis, decides the operation conditions of each part of the heating station by calculating the checking flow, the cold and hot loads and the generated energy, transmits the decision results to the load adjusting module, and uploads the operation parameters of the operation checking module and the prediction results of the learning prediction module to the cloud database, and the decision module provides decisions with a preset time step as a period.
The load adjusting module reads the decision result from the real-time decision module, compares the decided operation parameters with fault data of the cloud database, and transmits an adjusting signal to the DCS after confirming that the system can normally operate; if the operation cannot be performed normally after the adjustment, the adjustment signal is ignored to maintain the original working condition, and the error decision is uploaded to the cloud. The load adjusting module adjusts the load according to a preset time step length as a period.
The cloud database comprises a normal operation parameter database, a fault operation parameter database, a prediction result database and a fault decision database, and is used for storing normal operation parameters, adjustment decisions, fault operation parameters and corresponding decisions of the heating station. The cloud database transmits and reads corresponding data according to the requirements of other modules, and the data comparison function can compare the data provided by the other modules with the data in the database and find similar data. When the prediction result error and the fault decision in a period of time exceed the threshold value, the cloud database sends out a reminding signal to prompt maintenance personnel to update the learning prediction module.
The geothermal coupling solar heating station is divided into a heating mode and a cooling mode according to the operation mode. When the summer temperature is higher, the cooling load is larger, the geothermal energy system enters a refrigeration mode, the first electromagnetic valve 13 is closed, the second electromagnetic valve 14 is closed, the third electromagnetic valve 15 is closed, the fourth electromagnetic valve 16 is opened, high-temperature water flowing in from the user side enters the shallow buried pipe for cooling through the fourth water pump 4, then flows out from the shallow buried pipe group 7 and enters the closed cooling tower 8 for cooling, and the first water pump 1 conveys the low-temperature water to the user side. When the cooling circuit cannot meet the refrigeration load, the second electromagnetic valve 14 is opened, the photovoltaic power generation device 11 drives the heating device 9 to cool the underground water storage tank 6, and the low-temperature water is delivered to the user side by flowing into the cooling circuit through the third water pump 3. When the temperature in autumn and winter is lower, the heat load is larger, the geothermal energy system enters a heating mode, the first electromagnetic valve 13 is opened, the second electromagnetic valve 14 is closed, the third electromagnetic valve 15 is opened, the fourth electromagnetic valve 16 is closed, low-temperature water flowing in from the user side exchanges heat with the deep buried pipe 5 through the second water pump 2, the deep buried pipe 5 does not take heat, medium-deep geothermal energy is led out through a heat exchange medium, meanwhile, the low-temperature water at the user side flows into the shallow buried pipe group 7 through the fourth water pump 4, flows out after the heat is obtained, and the high-temperature water is conveyed to the user side through the first water pump 1. When the cooling circuit cannot meet the heating load, the second electromagnetic valve 14 is opened, the photovoltaic power generation device 11 drives the heating device 9 to heat the underground water storage tank 6, and the water flows into the heating circuit through the third water pump 3 to convey the high-temperature water to the user side.
The geothermal coupling solar heat supply station is powered by a photovoltaic power generation system, the photovoltaic system can participate in grid-connected power generation on the premise of meeting the power demand of the system, the power which cannot participate in grid-connected peak regulation is stored by using the energy storage battery 12, and when the energy storage battery 12 is full of electric energy, the electric energy which cannot be connected in a grid is used for driving the cooling and heating device to store energy by using the underground water storage tank 6. When the photovoltaic power generation device 11 is not supplied with power, the energy storage battery 12 supplies energy to the geothermal energy system. When the energy storage battery 12 is used up and the photovoltaic power generation device 11 is not powered sufficiently, the geothermal energy system is switched to an external power supply mode.
The operation checking module records current operation parameters including water temperature and flow rate of an inlet and an outlet of a user end, water temperature and flow rate of an inlet and an outlet of the deep buried pipe 5, water temperature and flow rate of a shallow buried pipe 7, water temperature and flow rate of an inlet and an outlet of a cooling tower, temperature and water storage condition of an underground water storage tank 6, power generation power of a photovoltaic array 10 and electric quantity storage condition of an energy storage battery 12 by taking a preset time step as a period. The module records the current operating parameters in the form of date, time, device and working condition. The module presets abnormal parameter values according to historical operation parameters of the cloud database, has a fault judging function, and after the current operation parameters are obtained, the module firstly judges whether the flow is conserved or not based on the current operation mode, whether the heat is conserved or not, whether the electric quantity is conserved or not, and performs fault judgment by comparing preset values. When possible abnormal data exist, the module transmits current operation parameters to the cloud database, the cloud database compares the data according to three adjacent time operation parameters and historical fault database operation parameters, determines a fault condition and actively alarms, generates a fault guiding scheme, and uploads the fault data and the fault position to the cloud database.
The learning prediction module has two functions of training and predicting, reads historical normal operation parameters of the heating station from the cloud and combines the imported corresponding historical meteorological conditions as a data set, and the data set is divided into a training set and a verification set. The module trains a predictive model based on an integrated predictive algorithm combining nearest neighbor nodes, random forests, and reinforcement learning, and optimizes the predictive model using a validation set. After the prediction model is obtained, the learning prediction module predicts the cold and hot load and the generating capacity of the photovoltaic system after a time step according to the current meteorological conditions, the module and the operation checking module act with the same frequency, the operation parameters transmitted by the operation checking module are read for prediction by taking the preset time step as a period, and the prediction result is transmitted to the real-time decision module.
The real-time decision module gives decision opinion to the whole system according to the predicted value given by the learning prediction module and the current load given by the operation checking module, and uploads the operation parameters of the operation checking module and the predicted result of the learning prediction module to the cloud database. Taking winter heating regulation as an example, the method specifically comprises the following steps:
s1: the real-time decision module calculates the flow q of each device of the system according to the current running condition given by the running checking module i ,i＝1,2,3,...,n；
Heat e provided by each device i ,i＝1,2,3,...,n；
Power consumption w of each device i ,i＝1,2,3,...,n；
Wherein the power generation is positive, the power consumption is negative, and the external power supply quantity is w e 。
S2: and the real-time decision module makes a decision according to the predicted value given by the learning prediction module, and when the difference between the predicted value of the thermal load of the user side, the predicted value of the power generation amount of the photovoltaic system and the operation parameter exceeds a set threshold value, a decision program is started. The adjustable part comprises parameters such as the opening degree of each electromagnetic valve, the distribution mode of the generated energy of a photovoltaic system, the inlet and outlet flow of the underground water storage tank and the like, and the following relation is satisfied in the adjustment process:
wherein Q is the user side flow and E is the user side heat requirement.
S3: the real-time decision module transmits the device and the decision to be regulated to the load regulation module, and does not transmit signals to the parts which do not need to be regulated.
Preferably, the load adjusting module performs load adjustment according to the decision provided by the real-time decision module, the module firstly checks the adjusted working conditions, checks whether the working conditions of all parts after adjustment are normal, compares the adjusted working conditions with a fault operation parameter library in a cloud database and a fault decision database, confirms that the system can normally operate, transmits an adjusting signal to the DCS system, and uploads adjusting data to the cloud; if the system cannot normally run after being regulated, the regulating signal is ignored to maintain the original working condition, and the error decision is uploaded to the cloud.
Preferably, the cloud database includes a normal operation parameter library, a fault operation parameter library, a prediction result database, and a fault decision database, and the cloud database transmits and reads data to each module of the intelligent control system, specifically includes transmitting historical fault data to the operation checking module, storing the fault data of the operation checking module, transmitting historical normal operation parameters to the learning prediction module, storing the normal operation parameters and the prediction data transmitted by the real-time decision module, and storing the decision data of the load adjusting module. The cloud data has a data comparison function, and specifically comprises the steps of comparing fault data transmitted by the operation checking module with historical fault data, and comparing decision data of the load adjusting module with a fault database and a fault decision database. When the prediction result error and the fault decision in a period of time exceed the threshold value, the cloud database sends out a reminding signal to prompt maintenance personnel to update the learning prediction module.
Preferably, as shown in fig. 3, the training of the prediction model in the learning prediction module includes the following steps:
step S1: and acquiring the operation parameters of the system in normal operation through the cloud database, importing the corresponding historical climate conditions, and carrying out normalization processing on the operation parameters and the historical climate conditions.
Step S2: and analyzing main factors influencing the heat load, the cold load and the power generation capacity of the photovoltaic system by using a correlation analysis method, and eliminating factors with low influence. And selecting the processed data as a data set.
Step S3: for the data sets, a cross validation method is used to divide the training set and the validation set respectively.
Step S4: and constructing an integrated algorithm prediction model according to the training set and the verification set. The integrated algorithm prediction model consists of three parts, namely a nearest neighbor algorithm (KNN), a Random Forest algorithm (Random Forest) and an adaptive lifting algorithm (AdaBoost), and the prediction result is weighted according to the prediction result of each part.
Step S5: the predictive model will back-propagate the weighting of the predictive results of the partial algorithms based on the deviation of the predicted data from the actual operating parameters. And retraining the predictive model when the predicted result and the operation parameter continuously deviate for a plurality of times within a certain time.
Step S6: repeating the steps S3-S5 to respectively obtain a user side heat demand prediction model, a user side cold demand prediction model and a photovoltaic system power generation amount prediction model.
Further, the step S1 specifically includes the following steps:
s1.1: the operating parameters in normal operation comprise 8 influence factors including selected time point, corresponding time point instant temperature, highest air temperature on the day, lowest air temperature on the day, sunlight time on the day, air humidity on the day, average wind speed on the day and date.
S1.2: normalizing the data:
wherein x is i,j Representing a j-th parameter in the i-th set of data; x's' i,j Represents x i,j Normalizing the obtained parameters; min (x) ,j ) Represents the minimum value, max (x ,j ) Representing the maximum value of the j-th parameter in each group, the user thermal load y i The same method is used for initialization.
Further, the step S2 specifically includes the following steps:
s2.1: according to the preset time step, a gray correlation method is used for respectively selecting the user heat load, the user cold load and the power generation amount of the photovoltaic array at the starting time point of each time step as a parent sequence, and the child sequence is used for selecting the time point, the instantaneous temperature at the corresponding time point, the highest air temperature on the day, the lowest air temperature on the day, the sun time on the day, the air humidity on the day, the average wind speed on the day, the date and the like.
S2.2: the temperature and the sunlight time are inversely related to the thermal load, and as the influencing factors increase, the thermal load of a user gradually decreases, and in the thermal load correlation analysis, the reciprocal operator needs to be multiplied:
X i 'D 1 ＝[x i ' ,1 d 1 ,x i ' ,2 d 2 ,...,x i ' ,j d j ]
wherein X is i ' represents the ith subsequence; x is x i ' ,j Representing a j-th parameter in the i-th set of data; d (D) 1 Is a reciprocal operator.
S2.3: and respectively obtaining gray correlation degrees of each influence factor on the user side heat requirement, the user side cold requirement and the photovoltaic array power generation capacity through a gray correlation method, and removing low correlation factors except for time points from the influence factors. And the processed data is used as a data set R, expressed as:
Further, the step S3 specifically includes the following steps:
s3.1: the data set R processed in step S2 is divided by using a cross-validation method, wherein 5-fold cross-validation is selected, and the data set R is divided into 5 parts, wherein 4 parts are used as training sets and 1 part is used as validation set.
S3.2: repeating the step S3.1 for 5 times, and selecting different training sets each time to obtain 20 training sets and 5 verification sets.
Further, the step S4 specifically includes the following steps:
s4.1: establishing a nearest neighbor algorithm prediction model, and initially setting a nearest neighbor parameter k=6, wherein the distance measurement adopts the following formula:
wherein d is Manhattan distance between two points in n-dimensional space, x' i,j Representing the j-th parameter in the i-th set of data.
S4.2: the gaussian function is selected as the weighting based on distance, and the mean square error and the average absolute error are used as the regression loss function, expressed as:
where a, b, c are parameters and x is the distance from a point in space to the true value.
Wherein y is i Is the true value of the i-th parameter,and n is the number of dimensions, and is the predicted value of the ith parameter.
S4.3: training the adjacent algorithm model based on the training set and the verification set obtained in the step S3, taking the minimized regression loss function as a target, optimizing the values a, b and c and the k values in the Gaussian function through a gradient descent algorithm, and using a KD tree method to accelerate training until the loss function reaches the requirement, so as to obtain the nearest adjacent algorithm model.
S4.4: and (3) establishing a random forest regression prediction model, using the data set R obtained in the step (S2), using a bagging method to put back and sample, generating N training subsets, and taking the data outside the bag which is not extracted as a verification set.
S4.5: dividing the feature tree by adopting a CART method, randomly selecting m influence factors (m is less than or equal to j) from j influence factors to serve as split feature values of current nodes of the feature tree, and selecting a mean square error and average absolute error minimization criterion to serve as split criteria of the feature tree, wherein the split criteria are expressed as follows:
wherein y is i Is the true value of the i-th parameter,and n is the number of dimensions, and is the predicted value of the ith parameter.
S4.6: for a prediction model of the heat load of the user side, selecting super parameters of a random forest model, and initially selecting N tree =400 feature trees, each feature tree having a maximum depth of50, initially selecting the characteristic value of each decision tree as m try =3, the prediction result is the average value of each tree output. And iteratively calculating the quantity of the optimized feature trees and the feature values of the decision trees with the minimum error as a target until the error is smaller than a set threshold value, and obtaining a random forest model.
S4.7: establishing a self-adaptive lifting algorithm prediction model, selecting a training set obtained in the step S3, selecting n groups of training samples from the training set, and endowing the training samples with initial weights of w 1i ＝1/n,i＝1,2,3,…,n；
With W (1) = (W) 11 ,w 12 ,w 13 ,…,w 1n ) Representing the initial weight of the sample, pick d=50 represents the number of learners.
S4.8 performing the iteration d=1, 2,3, …, D training the D-th weak learner H d (x) When using weak learner H d (x) The prediction validation set outputs regression error epsilon d Calculating the maximum error E of samples on the training set d And relative error epsilon di Calculating the weight alpha of the weak learner in the final learner d According to the weight alpha d The weight w (d+1) of the update sample is expressed as:
E d ＝max(|y i -H d (x i )|)
W(d+1)＝(w d|1,1 ,w d|1,2 ,w d|1,3 ,…,w d|1,n )
wherein y is i Is the true value of the i-th parameter.
S4.9: training D wheels to obtain D groups of weak learners H d (x) Combining the individual weak learners according to the weak learner weights results in a reinforcement learner h (x) and optimizing the weak learner number D using the verification set data, expressed as:
s4.10: and (3) weighting and combining the three trained models to obtain an integrated algorithm prediction model, wherein the weights of the initial three models are 1/3.
Further, the step S5 specifically includes the following steps:
s5.1: and optimizing the integrated algorithm prediction model. Inputting the verification set data obtained in the step S3, respectively obtaining and outputting a trained nearest neighbor algorithm prediction model, a random forest regression prediction model and an adaptive lifting algorithm prediction model, and taking the square error of each model prediction result and a true value as a loss function. The final prediction result is obtained by weighting the prediction results of all the parts, the initial weight is 1/3, and the weight proportion of all the parts is optimized by using historical data until the loss function J θ Less than a set threshold:
D＝w 1 d 1 +w 2 d 2 +w 3 d 3
/>
d is an output result of the integrated prediction model; d, d i ,w i The method comprises the steps of predicting a result and a weight by a nearest neighbor algorithm, a random forest algorithm and a self-adaptive lifting algorithm respectively;w' i is the updated weight; d' is a true value; k is a parameter controlling the weight update.
S5.2: and uploading the prediction result of the integrated prediction model to a cloud database each time, and when the error condition of the prediction result and the true value of the integrated prediction model exceeds a set value within a certain time step, sending a signal by the cloud database to remind maintenance personnel to update the learning prediction module, and repeating the steps S1-S5.1 to retrain the prediction model.
It is to be understood that the foregoing description is only a part of the embodiments of the present invention, and that the equivalent changes of the system described according to the present invention are included in the protection scope of the present invention. Those skilled in the art can substitute the described specific examples in a similar way without departing from the structure of the invention or exceeding the scope of the invention as defined by the claims, all falling within the scope of protection of the invention.
Claims (10)
1. An intelligent control method for a geothermal coupling solar heating station is characterized by comprising the following steps:
S1: monitoring the operation parameters of the geothermal coupling solar heating station according to a preset monitoring period, and performing fault diagnosis on the obtained operation parameters; if the operation parameters are abnormal, comparing the operation parameters with the historical data, and determining a fault solution according to the comparison result; if the operation parameters are normal, turning to S2;
s2: taking the historical normal operation parameters of the geothermal coupling solar heating station and the corresponding historical meteorological conditions as data sets, dividing the data sets into training sets and verification sets, and constructing an integrated algorithm prediction model; training an integrated algorithm prediction model by using an integrated prediction algorithm based on the combination of a nearest neighbor node algorithm, a random forest algorithm and a reinforcement learning algorithm in a training set, optimizing the integrated algorithm prediction model by using a verification set, and predicting the requirements of a user side and the power generation capacity condition of a photovoltaic system;
s3: according to the normal operation parameters obtained in the step S1 and the prediction results of the step S2, the operation conditions of all parts of the geothermal coupling solar heating station are decided by calculating the checking flow, the cold and hot loads and the generated energy;
s4: calculating the decided operation parameters according to the decision result obtained in the step S3, comparing the decided operation parameters with fault data, and transmitting the adjusting signals to a DCS (distributed control system) of the geothermal coupling solar heat supply station after confirming that the power supply system and the heat supply system of the geothermal coupling solar heat supply station can normally operate; if the power supply system and the heating system of the geothermal coupling solar heating station cannot normally operate after adjustment, the adjustment signal is ignored to maintain the original working condition, and error adjustment data are uploaded.
2. The intelligent control method of geothermal coupling solar heating station according to claim 1, wherein in S1, the operation parameters include a selected time point, an instantaneous temperature corresponding to the time point, a highest air temperature on the day, a lowest air temperature on the day, a solar time on the day, an air humidity on the day, an average wind speed on the day, and a date;
normalizing the operation parameters before fault diagnosis:
wherein x is i,j Representing a j-th parameter in the i-th set of data; x's' i,j Represents x i,j Normalizing the obtained parameters; min (x) ,j ) Represents the minimum value, max (x ,j ) Representing the maximum value of the j-th parameter in each group.
3. The intelligent control method for geothermal coupling solar heating station according to claim 1, wherein in S2, the training method for the prediction model comprises:
s2.1: according to the historical normal operation parameters of the geothermal coupling solar heating station, importing corresponding historical meteorological conditions, and carrying out normalization processing on the historical normal operation parameters and the historical meteorological conditions;
s2.2: analyzing main factors influencing heat load, cold load and generating capacity of a photovoltaic system by using a correlation analysis method, removing low-influence data, and taking the rest data as a data set;
S2.3: dividing the data set into a training set and a verification set by using a cross verification method;
s2.4: constructing an integrated algorithm prediction model according to the training set and the verification set, respectively training a nearest neighbor algorithm prediction model, a random forest algorithm prediction model and a self-adaptive lifting algorithm prediction model, and weighting and outputting a prediction result to a final prediction result;
s2.5: the integrated algorithm prediction model updates corresponding weights according to the prediction results of the nearest neighbor algorithm prediction model, the random forest algorithm prediction model and the self-adaptive lifting algorithm prediction model and the deviation back propagation of the actual operation parameters of the geothermal coupling solar heating station; retraining the integrated algorithm prediction model constructed in the step S2.4 when the prediction result and the operation parameter of the integrated algorithm prediction model continuously deviate for a plurality of times within a certain time;
s2.6: and repeating the steps S2.3-S2.5 to respectively obtain a user side heat demand prediction model, a user side cold demand prediction model and a photovoltaic system power generation amount prediction model.
4. The intelligent control method of geothermal coupling solar heating station according to claim 3, wherein S2.2 specifically comprises:
s2.2.1: according to a preset time step, respectively selecting a user thermal load, a user cold load and the power generation amount of the photovoltaic array at the starting time point of each time step as a master sequence by using a gray correlation method; the subsequence selects a time point, an instantaneous temperature at a corresponding time point, a highest air temperature on the day, a lowest air temperature on the day, a sunshine time on the day, an air humidity on the day, an average wind speed on the day and a date;
S2.2.2: according to the temperature and sunlight time which are inversely related to the heat load, multiplying the heat load by a reciprocal operator in the heat load correlation analysis:
X′ i D 1 ＝[x′ i,1 d 1 ,x′ i,2 d 2 ,...,x′ i,j d j ]
wherein X 'is' i Representing the ith subsequence; x's' i,j Representing a j-th parameter in the i-th set of data; d (D) 1 Is a reciprocal operator;
s2.2.3: gray correlation degrees of all influence factors on the user side heat requirement, the user side cold requirement and the photovoltaic array power generation amount are respectively obtained through a gray correlation method, and low correlation factors except for time points are removed from the influence factors; and the processed data is used as a data set R, expressed as:
5. the intelligent control method of geothermal coupling solar heating station according to claim 3, wherein S2.3 specifically comprises:
s2.3.1: dividing the data set R obtained by processing in the step S2.2 by using a cross validation method, selecting 5-fold cross validation, dividing the data set R into 5 parts, wherein 4 parts are used as training sets and 1 part is used as validation set;
s2.3.2: step S2.3.1 is repeated 5 times, and different training sets are selected each time, so that 20 training sets and 5 verification sets are obtained.
6. The intelligent control method of geothermal coupling solar heating station according to claim 3, wherein S2.4 specifically comprises:
s2.4.1: establishing a nearest neighbor algorithm prediction model, setting a super parameter k, selecting Manhattan distance as distance measurement, selecting mean square error and average absolute error as loss functions, selecting Gaussian functions as weights based on distance, and obtaining the nearest neighbor algorithm prediction model based on a training set and a verification set obtained in the step S2.3, taking a minimized loss function as a target, and using KD tree acceleration training; expressed as:
Wherein y is i Is the true value of the i-th parameter,the predicted value of the ith parameter, n is the number of dimensions;
wherein a, b, c are parameters, x is the distance from a point in space to a true value;
s2.4.2: establishing a random forest regression prediction model, and initially selecting N tree Setting the maximum depth of each feature tree to be 50, and initially selecting the feature value of each feature tree to be m try The method comprises the steps of carrying out a first treatment on the surface of the Selecting the data set R obtained in the step S2.2, and using a bagging method to carry out put-back sampling to generate N tree Training subsets, wherein the data outside the bags which are not extracted are used as verification sets; dividing the feature tree by CART method, and randomly selecting m from j influencing factors try A influencing factor (m) try J) as the splitting characteristic value of the current node of the characteristic tree, selecting a mean square error and average absolute error minimization criterion as the splitting standard of the characteristic tree, and predicting the result as the average value output by each tree; iteratively calculating the quantity of the optimized feature trees and the feature values of the decision trees with the minimum error as a target until the error is smaller than a set threshold value to obtain a random forest prediction model;
s2.4.3: establishing a self-adaptive lifting algorithm prediction model, selecting a training set obtained in the step S2.3, selecting n groups of training samples from the training set, and endowing the training samples with initial weights of w 1i ＝1/n，i＝1，2,3, …, n, with W (1) = (W) 11 ，w 12 ，w 13 ，…，w 1n ) Representing the initial weight of the sample, D representing the number of learners; performing the iteration d=1, 2,3, …, D, training the D-th weak learner H d (x) When using weak learner H d (x) The prediction validation set outputs regression error epsilon d Calculating the maximum error E of samples on the training set d And relative error epsilon di Calculating the weight alpha of the weak learner in the final learner d According to the weight alpha d Updating the weight w (d+1) of the sample; training D wheels to obtain D groups of weak learners H d (x) Combining the weak learners according to the weak learner weights to obtain an enhanced learner h (x), and optimizing the weak learner number D by using the verification set data; expressed as:
E d ＝max(|y i -H d (x i )|)
W(d+1)＝(w d|1，1 ，w d|1，2 ，w d|1，3 ，…，w d|1，n )
wherein y is i Is the true value of the i-th parameter.
7. The intelligent control method of geothermal coupling solar heating station according to claim 3, wherein S2.5 specifically comprises:
s2.5.1: optimizing the integrated algorithm prediction model; inputting verification set data obtained in the step S2.3, respectively obtaining output by using a trained nearest neighbor algorithm prediction model, a random forest regression prediction model and an adaptive lifting algorithm prediction model, taking the mean square error of an integrated algorithm model prediction result and a true value as a loss function, and optimizing the weight proportion of each part by using historical data until the loss function J is reached θ Less than a set threshold:
D＝w 1 d 1 +w 2 d 2 +w 3 d 3
d is an output result of the integrated algorithm prediction model; d, d i ,w i The method comprises the steps of predicting a result and a weight by a nearest neighbor algorithm, a random forest algorithm and a self-adaptive lifting algorithm respectively; w' i Is the updated weight; d' is a true value; k is a parameter for controlling the weight update;
s2.5.2: and uploading the prediction result of the integrated algorithm prediction model to a cloud database every time, and when the prediction result error or the fault decision in a period of time exceeds a threshold value, sending a signal by the cloud database to remind maintenance personnel to update the learning prediction module, and repeating the steps S2.1-S2.5.1 to retrain the prediction model.
8. An intelligent control system of a geothermal coupling solar heating station, comprising:
the operation checking module is used for monitoring the operation parameters of the geothermal coupling solar heating station according to a preset monitoring period and performing fault diagnosis on the obtained operation parameters; if the operation parameters are abnormal, uploading the operation parameters to a cloud database; if the operation parameters are normal, transmitting the operation parameters to a real-time decision module;
the learning prediction module reads historical normal operation parameters of the geothermal coupling solar heat supply station from the cloud database, combines corresponding historical meteorological conditions to serve as a data set, divides the data set into a training set and a verification set, and constructs an integrated algorithm prediction model; training an integrated algorithm prediction model by using an integrated prediction algorithm based on the combination of a nearest neighbor node algorithm, a random forest algorithm and a reinforcement learning algorithm and optimizing the integrated algorithm prediction model by using a verification set, predicting the conditions of the demand of a user side and the power generation capacity of a photovoltaic system, and transmitting a prediction result to a real-time decision module;
The real-time decision module is used for respectively reading the normal operation parameters from the operation checking module and the prediction results of the learning prediction module, deciding the operation conditions of each part of the geothermal coupling solar heat supply station by calculating the checking flow, the cooling and heating load and the generating capacity, transmitting the decision results to the load adjusting module, and uploading the normal operation parameters from the operation checking module and the prediction results of the learning prediction module to the cloud database;
the load adjusting module is used for reading the decision result from the real-time decision module, calculating the decided operation parameters, comparing the decided operation parameters with fault data in the cloud database, and confirming that the power supply system and the heat supply system of the geothermal coupling solar heat supply station can normally operate and then transmitting an adjusting signal to the DCS system of the geothermal coupling solar heat supply station; if the power supply system and the heating system of the geothermal coupling solar heating station cannot normally operate after adjustment, ignoring the adjustment signal to maintain the original working condition, and uploading error adjustment data to a cloud database;
the cloud database is used for storing normal operation parameters, fault operation parameters, prediction result data and fault decision data and comparing the data with uploaded data.
9. The geothermal coupled solar heating plant intelligent control system of claim 8, wherein a monitoring period of the operation check module, a decision period of the learning prediction module, a decision period of the real-time decision module, and an adjustment period of the load adjustment module are consistent with a preset time step.
10. The intelligent control system of the geothermal coupling solar heating station according to claim 8, wherein the operation checking module monitors the operation parameters of the geothermal coupling solar heating station including the water temperature and the flow rate of the user-side inlet and outlet, the water temperature and the flow rate of the deep buried pipe inlet and outlet, the water temperature and the flow rate of the shallow buried pipe inlet and outlet, the water temperature and the flow rate of the cooling tower inlet and outlet, the temperature and the water storage condition of the underground water storage tank, the power consumption of the geothermal energy system, the power generation power of the photovoltaic array and the power storage condition of the energy storage battery, and the operation checking module records the current operation parameters in the form of date, time period, device and working condition; after the operation checking module presets an abnormal parameter range to obtain the current operation parameter, firstly judging whether the flow, the energy and the electric quantity of a power supply system and a heat supply system of the geothermal coupling solar heat supply station are conservation or not based on the current operation mode, and carrying out fault judgment by comparing the preset abnormal parameter range; when possible abnormal data exist, the operation checking module transmits current fault data to the cloud database, the cloud database compares the data according to the operation parameters of three adjacent time steps and the operation parameters of the historical fault database, determines a fault position, actively alarms, generates a fault guiding scheme, and uploads the fault data and the fault position to the cloud database;
The cloud database comprises a normal operation parameter database, a fault operation parameter database, a prediction result database and a fault decision database; the cloud database has a data comparison function, and specifically comprises the steps of comparing possible fault data transmitted by the operation checking module with historical fault data, comparing the decided operation parameters calculated by the load adjusting module with the fault operation database, and comparing a prediction result with corresponding real data; when the prediction result error or the fault decision exceeds a threshold value within a period of time, the cloud database sends out a reminding signal to prompt maintenance personnel to update the learning prediction module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310451141.4A CN116499023A (en) | 2023-04-24 | 2023-04-24 | Intelligent control method and system for geothermal coupling solar heating station |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310451141.4A CN116499023A (en) | 2023-04-24 | 2023-04-24 | Intelligent control method and system for geothermal coupling solar heating station |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116499023A true CN116499023A (en) | 2023-07-28 |
Family
ID=87329749
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310451141.4A Pending CN116499023A (en) | 2023-04-24 | 2023-04-24 | Intelligent control method and system for geothermal coupling solar heating station |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116499023A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116864190A (en) * | 2023-07-31 | 2023-10-10 | 江苏恒辉电气有限公司 | Flame-retardant fire-resistant anti-interference radiation-resistant control cable and detection method |
-
2023
- 2023-04-24 CN CN202310451141.4A patent/CN116499023A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116864190A (en) * | 2023-07-31 | 2023-10-10 | 江苏恒辉电气有限公司 | Flame-retardant fire-resistant anti-interference radiation-resistant control cable and detection method |
| CN116864190B (en) * | 2023-07-31 | 2024-01-30 | 江苏恒辉电气有限公司 | Flame-retardant fire-resistant anti-interference radiation-resistant control cable and detection method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108205717A (en) | A kind of photovoltaic generation power Multiple Time Scales Forecasting Methodology | |
| CN110909911B (en) | Aggregation method of multidimensional time series data considering space-time correlation | |
| CN106529719A (en) | Method of predicting wind power of wind speed fusion based on particle swarm optimization algorithm | |
| CN104810860A (en) | Power distribution method and power distribution device in wind power plant | |
| CN110752626A (en) | Rolling optimization scheduling method for active power distribution network | |
| CN110400018B (en) | Operation control method, system and device for coal-fired power plant pulverizing system | |
| CN116499023A (en) | Intelligent control method and system for geothermal coupling solar heating station | |
| CN112927097A (en) | Photovoltaic power generation short-term prediction method based on GRA-ABC-Elman model | |
| CN109376426B (en) | Wind power grid-connected power scheduling method and device | |
| CN113937825A (en) | DG double-layer optimization configuration method based on E-C-Kmeans clustering and SOP optimization | |
| CN113469412A (en) | Real-time operation strategy optimization method and system for comprehensive energy system | |
| CN116109095A (en) | Day-ahead optimal scheduling method and system for heating system considering supply and demand coordination | |
| CN117674216B (en) | Steam energy storage device group regulation and control method based on simulated learning mechanism | |
| CN110991703B (en) | Wind-solar-water combined power generation scheduling method based on improved NSGA-II algorithm | |
| CN108321801A (en) | Method and system for making day-ahead power generation plan of energy base system | |
| CN116227704A (en) | Edge calculation and cloud calculation-based heat supply energy consumption diagnosis and energy-saving regulation and control method | |
| CN116205148A (en) | Data center cooling load prediction method based on Wide & Deep model | |
| CN115034293B (en) | Dynamic reconstruction method of power distribution network based on improved double-scale spectral clustering algorithm | |
| CN116154875A (en) | Photovoltaic power station active power optimization distribution method based on TCN and error function | |
| CN112668764B (en) | Offshore wind farm energy storage system optimization configuration method based on cloud model and FCM algorithm | |
| CN115630331A (en) | Air-conditioning refrigeration station global energy-saving optimization and regulation and control method based on mathematical physical model | |
| CN115271244A (en) | Two-stage distribution robust optimization-based short-term peak regulation model of cascade hydropower station | |
| CN113159380B (en) | Comprehensive energy system operation optimization method considering demand response | |
| Oonsivilai et al. | Gas turbine optimal PID tuning by genetic algorithm using MSE | |
| CN111340267A (en) | Optimization control method and equipment of regional comprehensive energy system based on block data |
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 |