CN113203117A - Composite heating control method based on BP neural network - Google Patents

Composite heating control method based on BP neural network Download PDF

Info

Publication number
CN113203117A
CN113203117A CN202110354556.0A CN202110354556A CN113203117A CN 113203117 A CN113203117 A CN 113203117A CN 202110354556 A CN202110354556 A CN 202110354556A CN 113203117 A CN113203117 A CN 113203117A
Authority
CN
China
Prior art keywords
water tank
heating
load
time
day
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.)
Granted
Application number
CN202110354556.0A
Other languages
Chinese (zh)
Other versions
CN113203117B (en
Inventor
陈振乾
杨震
许波
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN202110354556.0A priority Critical patent/CN113203117B/en
Publication of CN113203117A publication Critical patent/CN113203117A/en
Application granted granted Critical
Publication of CN113203117B publication Critical patent/CN113203117B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

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
    • F24D19/1045—Arrangement or mounting of control or safety devices for water heating systems for central heating the system uses a heat pump and solar energy
    • 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
    • F24D2200/00—Heat sources or energy sources
    • F24D2200/11—Geothermal energy
    • 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
    • F24D2200/00—Heat sources or energy sources
    • F24D2200/12—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
    • F24D2200/00—Heat sources or energy sources
    • F24D2200/14—Solar energy

Abstract

The invention provides a composite heating control method based on a BP neural network, which is applied to a solar energy and soil source heat pump composite system and comprises the following steps: monitoring the outlet temperature of the load side of the water tank, and if the outlet temperature of the load side of the water tank does not meet the direct supply requirement of the water tank, adopting a ground source heat pump heating mode; if the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank, judging that the water tank has 0.1Qh,nom≤Qh≤AQh,nomAnd if so, adopting a water tank direct supply mode, otherwise, adopting a ground source heat pump heating mode. The invention discloses a BP neural network-based composite heating control method, which optimizes the starting conditions of a water tank direct supply mode and a soil source heat pump heating mode to enable a soil source heat pump system to be in a high-efficiency state, thereby improving the energy efficiency of a unit.

Description

Composite heating control method based on BP neural network
Technical Field
The invention belongs to the technical field of ground source heat pump air conditioners, and particularly relates to a composite heating control method based on a BP neural network.
Background
In cold regions, the solar energy-soil source heat pump composite system can solve the problem of soil heat imbalance caused by the operation of a single ground source heat pump, and has the advantages of high efficiency, environmental protection and energy conservation, so that the solar energy-soil source heat pump composite system is widely applied. Common heating strategies of a composite system are divided into series connection, parallel connection and series-parallel connection according to the hot water using mode of a heat storage water tank in a solar heat collection system. At present, most scholars research the way to utilize heat in the hot water storage tank more efficiently on the basis of series, parallel and parallel connection, so as to improve the operation energy efficiency of the system. For example, the outlet water temperature on the load side of the water tank is divided into 4 sections, and different water tank hot water utilization modes are designed for different temperature sections.
The existing widely-adopted mode of directly supplying heat to the tail end by using hot water in a water tank is that when the temperature of the water tank meets the starting temperature required by direct heat supply, the operation of the soil source heat pump unit is stopped. Due to the lack of dynamic characteristics considering the actual operation of the unit, the water tank direct supply mode is started when the heating load is large, and the water tank direct supply mode is switched to the heat pump heating condition when the load is small, so that the heat pump operates under low load for a long time, and the unit operation energy efficiency is low. And the direct supply mode of the water tank is started under a large load, and the heat supplementing rate of the solar heat collector to the water tank cannot enable the temperature of the water tank to be maintained above the starting temperature, so that the problem that the unit is frequently started and stopped is caused. And in the aspect of the operation and control method of the solar energy-soil source heat pump composite system, the research related to controlling the heat storage water tank to directly heat is less.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the method optimizes the starting conditions of a direct water tank supply mode and a ground source heat pump heating mode, enables a ground source heat pump system to be in a high-efficiency state, and accordingly improves the unit energy efficiency.
In order to solve the technical problem, an embodiment of the present invention provides a composite heating control method based on a BP neural network, which is applied to a solar energy and soil source heat pump composite system, and the composite heating control method includes:
monitoring the outlet temperature of the load side of the water tank, and if the outlet temperature of the load side of the water tank does not meet the direct supply requirement of the water tank, adopting a ground source heat pump heating mode; if the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank, judging whether the formula (1) is met, if so, adopting a direct supply mode of the water tank, otherwise, adopting a heating mode of a soil source heat pump;
0.1Qh,nom≤Qh≤AQh,nomformula (1)
In the formula, Qh,nomRepresents the rated heating capacity, Q, of the soil source heat pumphIndicates the current actual heating load, and a indicates the control coefficient.
As a further improvement of the embodiment of the present invention, the method further includes:
and when the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first time, the control coefficient A is obtained.
As a further improvement of the embodiment of the present invention, the obtaining of the control coefficient a specifically includes:
step 10) predicting to obtain a hourly solar radiation value and a hourly heating load value;
step 11), calculating the direct heat supply quantity of the water tank by using the formula (2):
Figure BDA0002996791370000021
B＝cw,pρV
C＝AtηcηL
in the formula, QzgIndicating the direct heat supply of the water tank, ItRepresenting the predicted solar radiation value at time t, t0The moment t representing that the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first times,endIndicating a preset solar radiation end time, cw,pRepresents the specific heat capacity of water, ρ represents the density of water, V represents the tank volume, AtRepresenting collector area, ηcExpressing heat collecting efficiency, ηLRepresenting the heat loss rate of the water tank and the pipeline; t issIndicating the tank temperature, T, at the beginning of the tank direct feed modeeIndicating the temperature of the water tank at the end of the direct supply mode of the water tank;
step 12) calculating the sum of the heating loads in the interval of the partial load rate of 0.1-m by using the formula (3):
Figure BDA0002996791370000022
in the formula, Qh,mRepresents the sum of heating loads in a section of a partial load factor of 0.1 to m, t0Indicating the temperature of the outlet on the load side of the tankMoment t of first meeting the direct supply requirement of the water tankh,endIndicating the end of the preset heating load, Qh,tRepresents the predicted heating load value, Q, at time th,nomRepresenting the rated heating capacity of the soil source heat pump; m is 0.2+ q × Δ p, q represents the number of times of returning to step 12) in step 13), the initial value of q is 0, and Δ p represents the preset coincidence interval;
step 13) of judging Qzg≤Qh,mAnd if yes, changing A to m, otherwise, changing A to 1 if m is equal to the preset maximum load rate, and otherwise, returning to the step 12).
As a further improvement of the embodiment of the present invention, the step 10) specifically includes:
step 101) acquiring a hourly solar radiation value and a hourly heating load value of n days before a prediction day, the highest ambient temperature and the lowest ambient temperature of the n days before the prediction day, and the highest ambient temperature and the lowest ambient temperature of the prediction day; n represents an integer of 2 or more;
step 102) establishing a solar radiation amount prediction model based on the BP neural network and a building load prediction model based on the BP neural network by using the data collected in the step 101), predicting by using the solar radiation amount prediction model to obtain a hourly solar radiation value of a predicted day, and predicting by using the building load prediction model to obtain a hourly heating load value of the predicted day.
As a further improvement of the embodiment of the invention, in the solar radiation amount prediction model based on the BP neural network, the input parameters are a time-by-time solar radiation value of a preset radiation time period n days before a prediction day, a maximum ambient temperature of the prediction day, and an average value of the maximum ambient temperature of the prediction day and the minimum ambient temperature of the prediction day, the BP neural network adopts a Levenberg-marrdquat algorithm, the number of input nodes is 24, the number of nodes of a hidden layer is 12, a tan sig form is selected for a neuron excitation function of the hidden layer, a logsig form is adopted for an excitation function of an output layer, and the time-by-time solar radiation value of the prediction day is output.
As a further improvement of the embodiment of the present invention, in the building load prediction model based on the BP neural network, the input parameters are a time-by-time load value of a preset heating time period n days before a prediction day, a highest ambient temperature and a lowest ambient temperature of the previous n days, a highest ambient temperature and a lowest ambient temperature of the prediction day, and a difference value between the prediction day and the highest ambient temperature of the previous day, the BP neural network adopts a Levenberg-Marquardt algorithm, the number of input nodes is 33, the number of hidden layer nodes is 14, the hidden layer neuron excitation function selects a tan sig form, the output layer excitation function adopts a logsig form, and the time-by-time heating load value of the prediction day is output.
Compared with the prior art, the technical scheme of the invention has the following beneficial effects: according to the load heating control method based on the BP neural network, provided by the embodiment of the invention, when the heating load is small and the outlet temperature of the load side of the water tank meets the direct heating requirement, the operation of the heat pump unit can be stopped, and the direct supply mode of the water tank is started. The problem that the heat pump unit runs at a low load rate to cause low efficiency of the heat pump unit can be avoided, and the problem that the heat pump unit is frequently started and stopped because the solar radiation amount cannot maintain the outlet temperature of the load side of the water tank for a long time above the direct supply starting temperature of the water tank due to overlarge heating load can also be avoided.
Drawings
FIG. 1 is a flow chart of a control method of an embodiment of the present invention;
FIG. 2 is a schematic diagram of a model for hourly solar radiation prediction in a method according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a model for time-to-time heating load prediction in a method according to an embodiment of the invention;
fig. 4 is a diagram of a simulation result of a start-stop condition of a unit in an embodiment of a method and an existing water tank direct supply mode, where fig. 4(a) is a diagram of a simulation of a start-stop condition of a unit in an existing water tank direct supply mode, and fig. 4(b) is a diagram of a simulation of a start-stop condition of a unit in an embodiment of a method in the embodiment of the invention;
FIG. 5 is a diagram of the COP simulation results of the unit in the method of the embodiment of the present invention and the conventional water tank direct supply mode.
Detailed Description
The technical solution in the embodiments of the present invention will be described more clearly and completely with reference to the accompanying drawings in the embodiments of the present invention.
The embodiment of the invention provides a BP neural network-based composite heating control method applied to a solar energy and soil source heat pump composite system, which comprises the following steps of:
the solar energy and soil source heat pump composite system operates, the outlet temperature of the load side of the water tank is monitored, and if the outlet temperature of the load side of the water tank does not meet the direct supply requirement of the water tank, a soil source heat pump heating mode is adopted; if the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank, judging whether the formula (1) is met, if so, adopting a direct supply mode of the water tank, otherwise, adopting a heating mode of a soil source heat pump.
0.1Qh,nom≤Qh≤AQh,nomFormula (1)
In the formula, Qh,nomRepresents the rated heating capacity, Q, of the soil source heat pumphIndicates the current actual heating load, and a indicates the control coefficient.
The water tank load side outlet temperature meeting the water tank direct supply requirement means that the water tank load side outlet temperature is larger than or equal to a preset temperature value, and the preset temperature value is generally set according to the heat exchange form at the tail end of a building and the indoor temperature requirement.
According to the load heating control method based on the BP neural network, when the heating load is small and the outlet temperature of the load side of the water tank meets the direct heating requirement, the operation of the heat pump unit can be stopped, the direct supply mode of the water tank is started, the problem that the efficiency of the heat pump unit is low due to the fact that the heat pump unit operates at a small load rate can be solved, and the problem that the heat pump unit is frequently started and stopped due to the fact that the heating load is too large and the solar radiation amount cannot maintain the outlet temperature of the load side of the water tank to be higher than the direct supply starting temperature of the water tank for a long time can be solved.
Preferably, in the method according to the embodiment of the present invention, the operation cycle of the solar energy and soil source heat pump combined system is 24 hours, and one operation cycle refers to a cycle from 0: 00 to 24: 00. the control coefficient a is updated once per operating period, i.e. there is a corresponding control coefficient per day.
Preferably, when the outlet temperature of the load side of the water tank on the day first meets the direct supply requirement of the water tank, the control coefficient A of the day is obtained. In subsequent system operation, the tank is satisfied each timeWhen required, the signal quality is judged to be 0.1Qh,nom≤Qh≤AQh,nomAnd determining whether to start the direct water supply mode.
Further, the obtaining of the control coefficient a specifically includes:
step 10) predicting to obtain a hourly solar radiation value and a hourly heating load value of the same day;
step 11), calculating the direct heat supply quantity of the water tank by using the formula (2):
Figure BDA0002996791370000041
B＝cw,pρV
C＝AtηcηL
in the formula, QzgIndicating the direct heat supply of the water tank, ItRepresenting the predicted solar radiation value at time t, t0The moment t representing that the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first times,endIndicating a preset solar radiation end time, cw,pRepresents the specific heat capacity of water, ρ represents the density of water, V represents the tank volume, AtRepresenting collector area, ηcExpressing heat collecting efficiency, ηLRepresenting the heat loss rate of the water tank and the pipeline; t issIndicating the tank temperature, T, at the beginning of the tank direct feed modeeIndicating the tank temperature at the end of the tank direct feed mode.
Step 12) calculating the sum of the heating loads in the interval of the partial load rate of 0.1-m by using the formula (3):
Figure BDA0002996791370000051
in the formula, Qh,mRepresents the sum of heating loads in a section of a partial load factor of 0.1 to m, t0The moment t representing that the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first timeh,endIndicating the end of the preset heating load, Qh,tRepresents the predicted heating load value, Q, at time th,nomRepresenting the rated heating capacity of the soil source heat pump; m is 0.2+ q × Δ p, q represents stepStep 13) returning to the step 12), wherein the initial value of q is 0, and Δ p represents a preset load rate interval;
step 13) of judging Qzg≤Qh,mAnd if yes, changing A to m, otherwise, if m is equal to the preset maximum load rate, changing A to 1, otherwise, returning to the step 12).
Wherein the preset duty interval is preferably 0.1. If the interval is too small, errors of solar radiation and load prediction can be amplified, and the effectiveness of the control method is reduced; if the interval is selected too large, the control effect of the control coefficient A is not obvious. The preset maximum load rate is preferably 0.5, and generally, the heat pump unit is in an inefficient operation state when the load rate is less than 0.5, and the operation efficiency is better when the load rate is more than 0.5. The preset maximum load rate also needs to be determined according to the actual technical parameters of the heat pump unit. Since the lower the load factor of the heat pump unit, the lower the operating efficiency, e.g., the efficiency at the load factor of 0.2 is lower than the efficiency at 0.3, the method of this embodiment sequentially changes from 0.1 to 0.5 when obtaining the control coefficient a, thereby maximizing the effect of the control method.
In order to speed up the speed of obtaining the control coefficient and make a mode control strategy in time, preferably, step 10) may be executed when the previous operation cycle of the solar energy and soil source heat pump combined system is ended and a new operation cycle is started.
Further, the step 10) specifically includes:
step 101) acquiring a hourly solar radiation value and a hourly heating load value of the previous n days of the day, the highest ambient temperature and the lowest ambient temperature of the previous n days, and the highest ambient temperature and the lowest ambient temperature of the day; n represents an integer of 2 or more; wherein the highest ambient temperature and the lowest ambient temperature of the day are available from a weather forecast system;
and 102) establishing a solar radiation amount prediction model based on the BP neural network and a building load prediction model based on the BP neural network by using the data collected in the step 101), predicting to obtain a hourly solar radiation value of the day by using the solar radiation amount prediction model, and predicting to obtain a hourly heating load value of the day by using the building load prediction model.
In view of obtaining a solar radiation amount prediction model based on the BP neural network and a building load prediction model based on the BP neural network with high accuracy, n is preferably 2.
The solar radiation amount prediction model based on the BP neural network is shown in FIG. 2, input parameters of the model are a time-by-time solar radiation value of a preset radiation time period n days before the current day, the highest environment temperature of the current day and an average value of the highest environment temperature and the lowest environment temperature of the current day, the BP neural network adopts a Levenberg-Marquardt algorithm, the number of input nodes is 24, the number of nodes of a hidden layer is 12, a tan sig form is selected for a neuron excitation function of the hidden layer, a logsig form is adopted for an output layer excitation function, and the time-by-time solar radiation value of the current day is output.
The building load prediction model based on the BP neural network is shown in FIG. 3, the input parameters are a time-by-time load value of a preset heating time period n days before the current day, the highest ambient temperature and the lowest ambient temperature of the current day, and a difference value between the highest ambient temperature of the current day and the highest ambient temperature of the previous day, the BP neural network adopts a Levenberg-Marquardt algorithm, the number of input nodes is 33, the number of hidden layer nodes is 14, a tan sig form is selected for a hidden layer neuron excitation function, a logsig form is adopted for an output layer excitation function, and the time-by-time heating load value of the current day is output.
An embodiment is provided below, in which the heating floor is a sinking sun, and the heating time is set to 8: 00-20: 00, solar radiation time of 6: 00-16: 00. the hybrid heating control method of the present embodiment is as follows:
step 100) before the operation cycle of the prediction day starts or when the operation cycle of the prediction day starts, the solar energy and soil source heat pump combined system predicts a hourly solar radiation value and a hourly heating load value of the prediction day, and specifically comprises the following steps:
step 1001) collecting historical data such as building load, solar radiation amount, environmental temperature and the like, wherein the historical data comprises the following steps of collecting and predicting 6 days before and two days before the day: 00-16: the time-by-time solar radiation value at 00 and 8: 00-20: a hourly load value at 00 hours, a maximum ambient temperature and a minimum ambient temperature on the first two days, the previous day, and the predicted day.
Step 1002) establishing a solar energy time-by-time radiation prediction model based on a BP neural network, and inputting the prediction day-ahead and two days-ahead 6: 00-16: the hourly solar radiation value at 00 hours, the maximum environmental temperature of the predicted day, and the average value of the maximum and minimum environmental temperatures of the predicted day are used as parameters, a Levenberg-Marquardt algorithm is adopted, the number of input nodes is 24, the number of nodes of a hidden layer is 12, a hidden layer neuron excitation function selects a tansig form, an output layer excitation function adopts a logsig form, and the hourly solar radiation value of the predicted day is output.
Step 1003) establishing a time-by-time load prediction model based on the BP neural network, and inputting the prediction day-ahead and two days-ahead 8: 00-20: the method comprises the steps of taking a hourly load value at 00 hours, the highest and lowest ambient temperatures of the previous two days, the previous day and the prediction day, and the difference value of the highest temperature of the prediction day and the previous day as parameters, adopting a Levenberg-Marquardt algorithm, inputting 33 nodes, 14 nodes of a hidden layer, selecting a tan sig form for a neuron excitation function of the hidden layer, adopting a logsig form for an output layer excitation function, and outputting a hourly heating load value of the prediction day.
And 2000) in the process of predicting the operation of the solar energy and soil source heat pump composite system, monitoring the outlet temperature of the load side of the water tank, and if the outlet temperature of the load side of the water tank is less than 45 ℃, adopting a soil source heat pump heating mode.
Step 2001) when the outlet temperature of the load side of the water tank is more than or equal to 45 ℃, recording the time as t0Calculating the time t by using the equation (3)0The sum Q of heating loads in the following partial load rate sections of 0.1 to m (m is 0.2, 0.3, 0.4, 0.5, respectively)h,mCalculating the direct heat supply quantity Q of the water tank by using the formula (2)zg。
Step 2002) m is equal to 0.2, and time t is compared0After partial load Qh,0.2And direct heat supply quantity QzgIf the judgment condition Q is satisfiedzg≤Qh,0.2If the control coefficient a is 0.2, the process proceeds to step 300), and if not, the process proceeds to step 2003).
Step 2003) m ═ m0.3, comparison time t0After partial load Qh,0.3And direct heat supply quantity QzgIf the judgment condition Q is satisfiedzg≤Qh,0.3If the control coefficient a is 0.3, the process proceeds to step 300), otherwise, the process proceeds to step 2004).
Step 2004) m is 0.4, and time t is compared0After partial load Qh,0.4And direct heat supply quantity QzgIf the judgment condition Q is satisfiedzg≤Qh,0.4If yes, the control coefficient a is 0.4, and the process proceeds to step 300), otherwise, the process proceeds to step 2005).
Step 2005) m is 0.5, and time t is compared0After partial load Qh,0.5And direct heat supply quantity QzgIf the judgment condition Q is satisfiedzg≤Qh,0.5If yes, the control coefficient a is 0.5, and the process proceeds to step 300), otherwise, the control value a is 1.
Step 300) monitoring the outlet temperature of the load side of the water tank in the process of predicting the heating operation of the daily composite system, and if the outlet temperature of the load side of the water tank does not meet the direct supply requirement of the water tank, adopting a ground source heat pump heating mode; if the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank, judging whether the formula (1) is met, if so, adopting a direct supply mode of the water tank, otherwise, adopting a heating mode of a soil source heat pump.
A comparison example is provided, and a traditional water tank direct supply method is adopted, namely a direct supply mode is adopted when the outlet temperature of the load side of the water tank is more than or equal to 45 ℃. A typical day of the heating season was selected to compare the operation of the examples of the present invention and the comparative examples, and the results are shown in fig. 4 and fig. 5. In fig. 4 and 5, strategy 1 is a conventional tank direct supply method of a hybrid system, and strategy 2 is a hybrid heating control method according to an embodiment of the present invention.
As shown in fig. 4, the operation is performed under the strategy 1, and the unit is started for 9 times in the whole heating time, especially for 8 times between 11:00 and 14: 00. Because the solar radiation is small in the period, the heat supplementing rate of the solar heat collector to the water tank cannot enable the outlet temperature of the load side of the water tank to be maintained at 45 ℃ or above, and the phenomenon of intermittent direct supply of the water tank occurs, so that the unit is started and stopped frequently. And when the heat pump operates under the strategy 2, the heat pump is stopped for 2 times, and the control value A of the day is 0.3, so that the direct supply mode of the water tank is started when the partial load rate is not more than 0.3 and is 12: 00-15: 00.
As shown in fig. 5, the average COP of the unit under the operation of the strategy 1 is 3.90 in the day, and the average COP of the unit under the operation of the strategy 2 is 4.02, which is improved by 3.12%. Compared with strategy 1, the unit is operated under a larger heating load under strategy 2 for a longer time, so that a higher COP is obtained.
The foregoing illustrates and describes the principles, general features, and advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are intended to further illustrate the principles of the invention, and that various changes and modifications may be made without departing from the spirit and scope of the invention, which is also intended to be covered by the appended claims. The scope of the invention is defined by the claims and their equivalents.

Claims (6)

1. A composite heating control method based on a BP neural network is characterized by being applied to a solar energy and soil source heat pump composite system, and comprises the following steps:
monitoring the outlet temperature of the load side of the water tank, and if the outlet temperature of the load side of the water tank does not meet the direct supply requirement of the water tank, adopting a ground source heat pump heating mode; if the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank, judging whether the formula (1) is met, if so, adopting a direct supply mode of the water tank, otherwise, adopting a heating mode of a soil source heat pump;
0.1Qh,nom≤Qh≤AQh,nomformula (1)
In the formula, Qh,nomRepresents the rated heating capacity, Q, of the soil source heat pumphIndicates the current actual heating load, and a indicates the control coefficient.
2. The composite heating control method based on the BP neural network according to claim 1, further comprising:
and when the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first time, the control coefficient A is obtained.
3. The hybrid heating control method based on the BP neural network according to claim 2, wherein the calculating the control coefficient a specifically includes:
step 10) predicting to obtain a hourly solar radiation value and a hourly heating load value;
step 11), calculating the direct heat supply quantity of the water tank by using the formula (2):
Figure FDA0002996791360000011
B＝cw,pρV
C＝AtηcηL
in the formula, QzgIndicating the direct heat supply of the water tank, ItRepresenting the predicted solar radiation value at time t, t0The moment t representing that the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first times,endIndicating a preset solar radiation end time, cw,pRepresents the specific heat capacity of water, ρ represents the density of water, V represents the tank volume, AtRepresenting collector area, ηcExpressing heat collecting efficiency, ηLRepresenting the heat loss rate of the water tank and the pipeline; t issIndicating the tank temperature, T, at the beginning of the tank direct feed modeeIndicating the temperature of the water tank at the end of the direct supply mode of the water tank;
step 12) calculating the sum of the heating loads in the interval of the partial load rate of 0.1-m by using the formula (3):
Figure FDA0002996791360000012
in the formula, Qh,mRepresents the sum of heating loads in a section of a partial load factor of 0.1 to m, t0The moment t representing that the outlet temperature of the load side of the water tank meets the direct supply requirement of the water tank for the first timeh,endIndicating the end of the preset heating load, Qh,tRepresents the predicted heating load value, Q, at time th,nomRepresenting the rated heating capacity of the soil source heat pump; m is 0.2+ q × Δ p, q represents the number of times of returning to step 12) in step 13), the initial value of q is 0, and Δ p represents the preset coincidence interval;
step 13) of judging Qzg≤Qh,mAnd if yes, changing A to m, otherwise, changing A to 1 if m is equal to the preset maximum load rate, and otherwise, returning to the step 12).
4. The hybrid heating control method based on the BP neural network according to claim 3, wherein the step 10) specifically comprises:
step 101) acquiring a hourly solar radiation value and a hourly heating load value of n days before a prediction day, the highest ambient temperature and the lowest ambient temperature of the n days before the prediction day, and the highest ambient temperature and the lowest ambient temperature of the prediction day; n represents an integer of 2 or more;
step 102) establishing a solar radiation amount prediction model based on the BP neural network and a building load prediction model based on the BP neural network by using the data collected in the step 101), predicting by using the solar radiation amount prediction model to obtain a hourly solar radiation value of a predicted day, and predicting by using the building load prediction model to obtain a hourly heating load value of the predicted day.
5. The hybrid heating control method according to claim 4, wherein the input parameters in the solar radiation amount prediction model based on the BP neural network are a time-by-time solar radiation value of a preset radiation time period n days before the predicted day, a maximum ambient temperature of the predicted day, and an average value of a maximum ambient temperature of the predicted day and a minimum ambient temperature of the predicted day, the BP neural network adopts a Levenberg-Marquardt algorithm, the number of input nodes is 24, the number of hidden layer nodes is 12, the hidden layer neuron excitation function selects a tan sig form, the output layer excitation function adopts a logsig form, and the output layer excitation function outputs the time-by-time solar radiation value of the predicted day.
6. The hybrid heating control method based on the BP neural network according to claim 4, wherein in the building load prediction model based on the BP neural network, the input parameters are a time-by-time load value of a preset heating time period n days before a prediction day, a maximum ambient temperature and a minimum ambient temperature of the previous n days, a maximum ambient temperature and a minimum ambient temperature of the prediction day, and a difference value between the prediction day and the maximum ambient temperature of the previous day, the BP neural network adopts a Levenberg-Marquardt algorithm, the number of input nodes is 33, the number of hidden layer nodes is 14, the hidden layer neuron excitation function selects a tansig form, the output layer excitation function adopts a logsig form, and the output is the time-by-time heating load value of the prediction day.
CN202110354556.0A 2021-03-29 2021-03-29 Composite heating control method based on BP neural network Active CN113203117B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110354556.0A CN113203117B (en) 2021-03-29 2021-03-29 Composite heating control method based on BP neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110354556.0A CN113203117B (en) 2021-03-29 2021-03-29 Composite heating control method based on BP neural network

Publications (2)

Publication Number Publication Date
CN113203117A true CN113203117A (en) 2021-08-03
CN113203117B CN113203117B (en) 2022-03-18

Family

ID=77026098

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110354556.0A Active CN113203117B (en) 2021-03-29 2021-03-29 Composite heating control method based on BP neural network

Country Status (1)

Country Link
CN (1) CN113203117B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN206582939U (en) * 2017-01-18 2017-10-24 西安汉清新能源科技有限公司 Water resource heat pump solar energy Combined heating system
CN107421163A (en) * 2017-04-13 2017-12-01 西安汉清新能源科技有限公司 Air electric energy twin-stage heating and refrigerating system and heating and refrigerating method
CN209263110U (en) * 2018-07-15 2019-08-16 甘肃快邦节能建筑有限公司 Solar heat pump combined heated device
CN110906571A (en) * 2019-11-30 2020-03-24 华南理工大学 Solar heat pump hot water system control strategy optimization method based on machine learning
CN112131715A (en) * 2020-09-01 2020-12-25 大连理工大学 TRNSYS-based long-short-term coupling heat storage solar heating system model and modeling method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN206582939U (en) * 2017-01-18 2017-10-24 西安汉清新能源科技有限公司 Water resource heat pump solar energy Combined heating system
CN107421163A (en) * 2017-04-13 2017-12-01 西安汉清新能源科技有限公司 Air electric energy twin-stage heating and refrigerating system and heating and refrigerating method
CN209263110U (en) * 2018-07-15 2019-08-16 甘肃快邦节能建筑有限公司 Solar heat pump combined heated device
CN110906571A (en) * 2019-11-30 2020-03-24 华南理工大学 Solar heat pump hot water system control strategy optimization method based on machine learning
CN112131715A (en) * 2020-09-01 2020-12-25 大连理工大学 TRNSYS-based long-short-term coupling heat storage solar heating system model and modeling method

Also Published As

Publication number Publication date
CN113203117B (en) 2022-03-18

Similar Documents

Publication Publication Date Title
CN109063925B (en) Optimized operation method for regional comprehensive energy system considering load aggregators
CN108964128B (en) Low-carbon economic dispatching solving method based on coordinated heat supply of electric boiler and heat storage device
CN109591659B (en) Intelligent learning pure electric vehicle energy management control method
CN101261021A (en) Multi- water tanks solar energy-heat pump composite hot-water system and its operation control method
CN105320118A (en) Method for electric power demand response control of air conditioning systems based on cloud platform
JP6215753B2 (en) How to operate an energy supply system
CN109390981B (en) Combined operation control method for new energy participating in electric quantity balance of sending-end power grid unit
CN115409396A (en) Comprehensive energy system multi-time scale scheduling method based on double-layer rolling optimization
CN111639819B (en) Multi-stage optimization control method for comprehensive energy park
CN112797526A (en) Energy-saving industrial circulating water conveying technology with different water supply temperatures and different temperature differences
CN114183796A (en) Optimal scheduling method and device based on electric heating and central heating multi-energy complementary system
CN111382902A (en) Regional comprehensive energy system energy storage optimization configuration method based on operation benefit increment
CN112001523A (en) Comprehensive energy multi-scale optimization scheduling control method and system considering multiple energy storages
CN113203117B (en) Composite heating control method based on BP neural network
CN112880133B (en) Flexible energy utilization control method for building air conditioning system
CN116817357B (en) Heating control method and device based on heating station, building and household
CN113152595B (en) Variable-frequency constant-pressure water supply system and energy-saving control method thereof
CN114136021B (en) Solar energy-ground source heat pump system control method and system combined with resource prediction
CN115682370A (en) Photovoltaic ice storage air conditioner control strategy adjusting method, device and system
CN112577100B (en) Heating compensation method and device and readable storage medium
CN115147007A (en) Park low-carbon economic energy utilization method based on electricity-carbon information and NSGA-II
CN115310770A (en) Novel hybrid energy storage optimal configuration method and system considering carbon transaction mechanism
CN114358604A (en) Urban centralized heating system time-sharing and zone-dividing regulation and control method considering building heat storage characteristics
CN108590968A (en) Blower fan tower barrel hydrogen storage energy conserving system and method
CN113315135A (en) Low-carbon building optimal scheduling method, system, terminal and 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
GR01 Patent grant
GR01 Patent grant