CN110410942B - Energy-saving optimization control method and system for cold and heat source machine room - Google Patents
Energy-saving optimization control method and system for cold and heat source machine room Download PDFInfo
- Publication number
- CN110410942B CN110410942B CN201910691078.5A CN201910691078A CN110410942B CN 110410942 B CN110410942 B CN 110410942B CN 201910691078 A CN201910691078 A CN 201910691078A CN 110410942 B CN110410942 B CN 110410942B
- Authority
- CN
- China
- Prior art keywords
- water pump
- host
- parameter
- load
- predicted
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/30—Control or safety arrangements for purposes related to the operation of the system, e.g. for safety or monitoring
- F24F11/46—Improving electric energy efficiency or saving
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
- F24F11/63—Electronic processing
- F24F11/64—Electronic processing using pre-stored data
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2130/00—Control inputs relating to environmental factors not covered by group F24F2110/00
- F24F2130/10—Weather information or forecasts
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2140/00—Control inputs relating to system states
- F24F2140/50—Load
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2140/00—Control inputs relating to system states
- F24F2140/60—Energy consumption
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention provides a cold and heat source machine room energy-saving optimization control method and a system, wherein the method comprises the following steps: establishing a neural network load prediction model, a neural network host parameter model and a neural network water pump parameter model according to historical data of operation parameters of systems and equipment of a cold and heat source machine room stored in a database; acquiring outdoor weather prediction parameters at a prediction moment, and acquiring a predicted air conditioner load by using a neural network load prediction model; transmitting the predicted air conditioner load to a host, a freezing side water pump and a cooling side water pump to obtain a host constraint condition and a water pump constraint condition; obtaining a host energy consumption range and a water pump energy consumption range according to the neural network host parameter model and the neural network water pump parameter model; and determining output parameters of the host and the water pump after multiple iterative calculations. The invention guides the output of the refrigerating output of the cold and heat source machine room by the predicted load, and simultaneously controls the output parameters of the host, the water pump and the like by utilizing the energy-saving group control algorithm established by historical data, thereby achieving better energy-saving effect.
Description
Technical Field
The invention relates to the technical field of air conditioner energy-saving control, in particular to a cold and heat source machine room energy-saving optimization control method and system.
Background
In recent years, artificial intelligence has been developed in the fields of image processing, natural voice processing, robots, and the like, and more common applications include image recognition, artificial customer service, and smart homes linked with smart speakers. Applications in the industrial field, such as equipment failure diagnosis, automatic driving, etc., are also gradually developed. In the field of central air-conditioning, there are many cases for predicting the change of the air-conditioning load trend in the future day by using artificial intelligence means such as a neural network, but there are few cases for controlling the energy-saving control of the cold and heat source machine room of the central air-conditioning by using the artificial intelligence technology.
The equipment of the air conditioning system is selected according to the maximum load to deal with the highest temperature or the lowest temperature weather in a year, and the weather is rare and rare in the year, and the equipment runs at full load in rare cases, so the surplus of the equipment is large. According to statistics, the energy consumption of the air conditioning system accounts for about 40% of the energy consumption of the whole building, and accounts for a large amount, so that the air conditioning system has a large energy-saving space. Most of energy consumption of the central air conditioning system is concentrated in a cold and heat source machine room, so the cold and heat source machine room is the key point of energy conservation.
The development of the automatic control system of the cold and heat source machine room roughly comprises three stages, wherein the first stage realizes basic monitoring and control of equipment and automatic machine adding and subtracting functions of the equipment, the second stage adds an energy-saving strategy (such as expert control) and realizes system group control, and the third stage realizes centralized control and energy-saving group control modes of multiple items on the basis of big data and a cloud platform.
With the rapid development of computer technology, the invention provides a cold and heat source machine room energy-saving optimization control method and system.
Disclosure of Invention
The invention aims to provide a cold and heat source machine room energy-saving optimization control method and system, which can be used for realizing energy-saving control of the cold and heat source machine room by fully utilizing the value of the existing historical data, deeply excavating the parameter characteristics of equipment and combining an energy-saving operation theory and the equipment characteristics.
The technical scheme provided by the invention is as follows:
the invention provides an energy-saving optimization control method for a cold and heat source machine room, which comprises the following steps:
establishing historical data of operating parameters of systems and equipment of a cold and heat source storage machine room in a database;
taking the load parameters in the database as training data to perform off-line training, and establishing a neural network load prediction model;
taking the host parameters in the database as training data to perform off-line training, and establishing a neural network host parameter model;
taking the water pump parameters in the database as training data to perform off-line training, and establishing a neural network water pump parameter model;
acquiring outdoor weather prediction parameters at the prediction time, and acquiring predicted air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters;
transmitting the predicted air conditioner load to a host and a water pump to respectively obtain corresponding host constraint conditions and water pump constraint conditions, wherein the water pump comprises a freezing side water pump and a cooling side water pump;
obtaining a host energy consumption range according to the neural network host parameter model and the host constraint condition;
obtaining a water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
and determining the predicted output parameters of the host and the water pump under the energy-saving group control strategy of the system through repeated iterative calculation according to the energy consumption range of the host and the energy consumption range of the water pump.
Further, after determining the predicted output parameters of the host and the water pump under the system energy-saving group control strategy through multiple iterative computations according to the host energy consumption range and the water pump energy consumption range, the method further comprises the following steps:
acquiring parameter information of a predicted moment through a sensor;
calculating actual air conditioner load and actual output parameters of the host and the water pump according to the parameter information;
and uploading the predicted air conditioner load, the predicted output parameters, the actual air conditioner load and the actual output parameters to the database.
Further, obtaining an outdoor weather prediction parameter at a prediction time, and obtaining a predicted air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameter specifically includes:
acquiring outdoor weather prediction parameters at a prediction moment;
acquiring a preliminary prediction air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters;
acquiring a predicted air conditioner load and an actual load parameter of preset time before a predicted time in the database;
and correcting the preliminary predicted air-conditioning load according to the error between the predicted air-conditioning load and the actual load parameter at the preset time to obtain the predicted air-conditioning load at the predicted time.
Further, obtaining the host energy consumption range according to the neural network host parameter model and the host constraint condition specifically includes:
acquiring a primary host energy consumption range according to the neural network host parameter model and the host constraint condition;
obtaining a host output parameter in the predicted output parameters of the preset time before the predicted time and a host actual output parameter in the actual output parameters in the database;
and correcting the primary host energy consumption range according to the error between the host output parameter at the preset time and the actual host output parameter to obtain the host energy consumption range at the predicted moment.
Further, obtaining the energy consumption range of the water pump according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition specifically comprises:
acquiring a preliminary water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
acquiring a water pump output parameter in a prediction output parameter of preset time before the prediction time and a water pump actual output parameter in an actual output parameter in the database;
and correcting the preliminary water pump energy consumption range according to the error between the water pump output parameter at the preset time and the actual water pump output parameter to obtain the water pump energy consumption range at the predicted moment.
The invention also provides an energy-saving optimization control system for the cold and heat source machine room, which comprises the following components:
the database module is used for establishing historical data of operating parameters of systems and equipment of a cold and heat source storage machine room in a database;
the model module is used for performing off-line training by taking the load parameters in the database established by the database module as training data to establish a neural network load prediction model;
the model module is used for performing off-line training by taking the host parameters in the database established by the database module as training data to establish a neural network host parameter model;
the model module is used for performing off-line training by taking the water pump parameters in the database established by the database module as training data to establish a neural network water pump parameter model;
the load prediction module is used for acquiring outdoor meteorological prediction parameters at the prediction time and acquiring the predicted air conditioner load according to the neural network load prediction model established by the model module and the outdoor meteorological prediction parameters;
the load transfer module is used for transferring the predicted air conditioner load acquired by the load prediction module to a host and a water pump to respectively acquire corresponding host constraint conditions and water pump constraint conditions, and the water pump comprises a freezing side water pump and a cooling side water pump;
the host parameter analysis module is used for obtaining a host energy consumption range according to the neural network host parameter model established by the model module and the host constraint condition determined by the load transfer module;
the water pump parameter analysis module is used for obtaining a water pump energy consumption range according to the neural network water pump parameter model established by the model module and the water pump constraint condition determined by the load transfer module and the host computer constraint parameter;
and the output parameter analysis module is used for determining the predicted output parameters of the host and the water pump under the system energy-saving group control strategy through repeated iterative calculation according to the host energy consumption range determined by the host parameter analysis module and the water pump energy consumption range determined by the water pump parameter analysis module.
Further, the method also comprises the following steps:
the parameter information acquisition module acquires parameter information of the predicted moment through a sensor;
the actual parameter analysis module calculates actual air conditioner load and actual output parameters of the host and the water pump according to the parameter information acquired by the parameter information acquisition module;
and the processing module uploads the predicted air conditioner load obtained by the load prediction module, the predicted output parameter obtained by the output parameter analysis module, the obtained actual air conditioner load and the actual output parameter to the database established by the database module.
Further, the load prediction module specifically includes:
a weather parameter acquiring unit for acquiring outdoor weather prediction parameters at the prediction time;
the load prediction unit acquires a preliminary predicted air conditioner load according to the neural network load prediction model and the outdoor meteorological prediction parameters acquired by the meteorological parameter acquisition unit;
a load parameter acquiring unit for acquiring a predicted air conditioner load and an actual load parameter of a preset time before a predicted time in the database;
and the load correction unit is used for correcting the preliminary predicted air-conditioning load obtained by the load prediction unit according to the error between the predicted air-conditioning load at the preset time and the actual load parameter obtained by the load parameter acquisition unit to obtain the predicted air-conditioning load at the predicted time.
Further, the host parameter analysis module specifically includes:
the host energy consumption calculation unit is used for acquiring a primary host energy consumption range according to the neural network host parameter model and the host constraint condition;
the host parameter acquiring unit is used for acquiring host output parameters in the predicted output parameters of the preset time before the predicted time in the database and host actual output parameters in the actual output parameters;
and the host energy consumption correction unit is used for correcting the primary host energy consumption range obtained by the host energy consumption calculation unit according to the host output parameter of the preset time and the host actual output parameter error obtained by the host parameter obtaining unit to obtain the host energy consumption range at the predicted moment.
Further, the water pump parameter analysis module specifically includes:
the water pump energy consumption calculation unit is used for acquiring a preliminary water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
the water pump parameter acquisition unit is used for acquiring a water pump output parameter in the predicted output parameters of the preset time before the predicted time in the database and a water pump actual output parameter in the actual output parameters;
and the water pump energy consumption correction unit is used for correcting the preliminary water pump energy consumption range obtained by the water pump energy consumption calculation unit according to the error between the water pump output parameter of the preset time and the actual water pump output parameter obtained by the water pump parameter obtaining unit to obtain the water pump energy consumption range at the predicted moment.
The cold and heat source machine room energy-saving optimization control method and the system provided by the invention can bring at least one of the following beneficial effects:
1. in the invention, based on historical data, a neural network load prediction model is established, and the air conditioner load at each future moment is accurately predicted.
2. In the invention, based on historical data, a neural network host parameter model and a neural network water pump parameter model are established, and the parameter relation between the output load capacity and the consumed power of the equipment is accurately obtained.
3. In the invention, the actual characteristics of equipment and the system group control energy-saving theory are deeply combined, and the cold and heat source machine room is operated under the required load in the most energy-saving mode under the condition of ensuring the safe operation of the system and the equipment.
Drawings
The following describes the preferred embodiments in a clearly understandable manner with reference to the accompanying drawings, and further describes the above characteristics, technical features, advantages and implementation manners of the cold and heat source machine room energy saving optimization control method and system.
Fig. 1 is a flowchart of a first embodiment of an energy-saving optimization control method for a cold and heat source machine room according to the present invention;
fig. 2, fig. 3, fig. 4 and fig. 5 are flowcharts illustrating a second embodiment of an energy-saving optimization control method for a cold and heat source machine room according to the present invention;
fig. 6 is a schematic structural diagram of a third embodiment of an energy-saving optimization control system for a cold and heat source machine room according to the invention;
fig. 7 is a schematic structural diagram of a fourth embodiment of an energy-saving optimization control method and system for a cold and heat source machine room according to the present invention;
FIG. 8 is a schematic diagram of the power versus cooling trend of the main machine;
FIG. 9 is a schematic diagram of the energy consumption and the flow rate variation trend of the water pump;
FIG. 10 is a schematic diagram of the cooling water temperature versus host and cooling water pump cluster variation;
FIG. 11 is a schematic diagram of cooling water temperature versus host and chilled water pump cluster variation.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following description will explain specific embodiments of the present invention with reference to the drawings of the specification. It is obvious that the drawings in the following description are only some examples of the invention, from which other drawings and embodiments can be derived by a person skilled in the art without inventive effort.
For the sake of simplicity, only the parts relevant to the present invention are schematically shown in the drawings, and they do not represent the actual structure as a product. In addition, in order to make the drawings concise and understandable, components having the same structure or function in some of the drawings are only schematically illustrated or only labeled. In this document, "one" means not only "only one" but also a case of "more than one".
In a first embodiment of the present invention, as shown in fig. 1, a method for optimally controlling energy saving in a cold and heat source machine room includes:
s100, establishing historical data of operation parameters of systems and equipment in a cold and heat source storage machine room of a database;
specifically, the neural network model must be trained through a large amount of data, so that historical data of operating parameters of systems and equipment of a cold and heat source storage machine room in a database are established, recorded data comprise host operating parameters, water pump operating parameters, operating parameters of the systems (such as supply and return water temperature, pressure, flow and valve states), loads of the systems, time, meteorological parameters and the like, recorded data quantity is as much and detailed as possible, and the recorded data quantity at least comprises a complete heating period and a complete cooling period of the last year. The more data used for training, the better the training. Therefore, the above parameters are only used for illustration, and do not represent that the database in the present embodiment includes only the above data, and the present embodiment does not specifically limit the data included in the database, depending on the need of building the neural network model.
Based on a large amount of historical data of the operation of the cold and heat source machine room, a load prediction, a host, a refrigeration water pump and a cooling water pump are established, and offline training is carried out.
S200, taking the load parameters in the database as training data to perform off-line training, and establishing a neural network load prediction model;
specifically, the neural network load prediction model is used for predicting the load condition of a building at a certain future time, the load of the building is related to the outdoor meteorological environment, the building structure, the building internal equipment and personnel, the building structure is fixed and unchanged, the building internal equipment and the personnel usually have certain rules which can be followed, the rules are extracted in a neural network mode, and the load prediction is carried out by combining outdoor meteorological environment parameters.
Therefore, the load parameters in the database are used as training data to carry out off-line training, and a neural network load prediction model is established. For example, data such as historical time load and outdoor meteorological parameters of each system are extracted, training data are established, a load value is used as a label, other data are used as training samples, the number of hidden layers and the number of neurons are properly selected, a proper learning rate is set, offline training is carried out, and a neural network load prediction model is established.
S300, taking the host parameters in the database as training data to perform off-line training, and establishing a neural network host parameter model;
specifically, the neural network host parameter model is used for summarizing the implementation operating power and the load condition of the host, and the load which can be provided and the power which can be consumed by the host are different under different water inlet and outlet temperatures, so that multi-dimensional training needs to be performed by utilizing historical data.
Therefore, the host parameters in the database are used as training data to perform off-line training, and a neural network host parameter model is established. For example, the output load, power, temperature of the inlet and outlet water of an evaporator and a condenser and the like of the host are extracted from the database, the power is used as output, other parameters are used as input layers, corresponding samples and labels are established, and the number of proper hidden layers, the number of neurons, an activation function and a learning rate are selected for off-line calculation. Each host machine should be independently modeled, and for the heat pump host machines capable of being heated, cooled and heated, models under the working conditions of refrigeration and heating should be respectively modeled.
S400, taking the water pump parameters in the database as training data to perform off-line training, and establishing a neural network water pump parameter model;
specifically, the neural network water pump parameter model is used for summarizing the flow and power consumption conditions of water pumps in different numbers and different frequencies, the difference of the running conditions of one water pump and two water pumps is large, and the conditions of various states need to be extracted from historical data for training. The neural network model for each class of device is not generic.
Therefore, the water pump parameters in the database are used as training data to perform off-line training, a neural network water pump parameter model is established, and the water pump characteristic curve and the resistance characteristic of a pipe network are substantially extracted. The resistance characteristics of the pipe network are different under different numbers of water pumps and hosts, and the resistance coefficient of the pipe network is unchanged only under the conditions that the opening number of all equipment is unchanged and the opening degree of the valve is unchanged. Therefore, the input parameters are more, the water pump parameter model cannot be established by a single water pump, and a neural network model is established by taking a water pump group of a system as a whole.
The neural network models in steps S200-S400 may be processed simultaneously, and there is no substantial chronological order.
S500, obtaining outdoor weather prediction parameters at the prediction time, and obtaining predicted air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters;
specifically, outdoor weather prediction parameters at the prediction time are obtained through weather forecast or a self-erected small weather station and the like, and the corresponding predicted air-conditioning load at the prediction time is obtained through a neural network load prediction model.
S600, transmitting the predicted air conditioner load to a host and a water pump to respectively obtain corresponding host constraint conditions and water pump constraint conditions, wherein the water pump comprises a freezing side water pump and a cooling side water pump;
specifically, the obtained predicted air conditioning load is transmitted to the host, the freezing side water pump and the cooling side water pump, and corresponding host constraint conditions and water pump constraint conditions, namely, allowable ranges of various parameters of the host and the water pump, are obtained respectively.
S700, obtaining a host energy consumption range according to the neural network host parameter model and the host constraint condition;
s800, obtaining a water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
and S900, determining the predicted output parameters of the host and the water pump under the system energy-saving group control strategy through multiple iterative calculations according to the host energy consumption range and the water pump energy consumption range.
Specifically, a neural network host parameter model and a neural network water pump parameter model are called, energy consumption ranges are calculated according to corresponding host constraint conditions and water pump constraint conditions respectively, the host energy consumption range and the water pump energy consumption range are considered comprehensively, and parameters of the corresponding host and the water pump with the lowest comprehensive energy consumption of the whole system under the system energy-saving group control strategy are determined as output parameters.
In the embodiment, the output of the refrigerating capacity of the cold and heat source machine room is guided by predicting the load of the air conditioner, the real-time load is matched, the waste of the refrigerating capacity is avoided, meanwhile, the energy-saving group control algorithm established by historical data is used for controlling output parameters of a host, a water pump and the like, the COP of the system is maximized on the premise of meeting the load and the safe operation of equipment, and a better energy-saving effect is achieved.
A second embodiment of the present invention is an optimized embodiment of the first embodiment, as shown in fig. 2, fig. 3, fig. 4, and fig. 5, and compared with the first embodiment, the present embodiment has the main improvement that after determining the predicted output parameters of the host and the water pump under the system energy saving group control strategy through multiple iterative computations according to the host energy consumption range and the water pump energy consumption range, the method further includes:
s950, acquiring parameter information of a predicted moment through a sensor;
s960, calculating according to the parameter information to obtain the actual air conditioner load and the actual output parameters of the host and the water pump;
s970 uploads the predicted air conditioner load, the predicted output parameters, the actual air conditioner load and the actual output parameters to the database.
Specifically, when the time reaches the predicted time in the first embodiment, the sensor acquires parameter information of the predicted time, where the parameter information includes parameter information such as water temperatures of the host and the water pump, and then calculates according to the acquired parameter information to obtain actual output parameters of the host and the water pump.
And uploading the predicted air conditioner load, the predicted output parameters, the actual air conditioner load and the actual output parameters to a database for storage, and providing a correction basis for next prediction control. In addition, the predicted air-conditioning load and the predicted output parameters at the predicted time can be uploaded to the database for storage at the calculated first time, and only the time, the property and the like of all data need to be marked in the database without being uploaded with the actual air-conditioning load and the actual output parameters.
S500, obtaining outdoor weather prediction parameters at the prediction time, and obtaining the predicted air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters specifically comprises:
s510, obtaining outdoor weather prediction parameters at the prediction time;
s520, acquiring a preliminary prediction air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters;
s530, acquiring a predicted air conditioner load and an actual load parameter of preset time before the predicted time in the database;
s540, correcting the preliminary predicted air-conditioning load according to the error between the predicted air-conditioning load at the preset time and the actual load parameter to obtain the predicted air-conditioning load at the predicted time;
the step S700 of obtaining the host energy consumption range according to the neural network host parameter model and the host constraint condition specifically includes:
s710, acquiring a preliminary host energy consumption range according to the neural network host parameter model and the host constraint condition;
s720, acquiring a host output parameter in the predicted output parameters of the preset time before the predicted time and a host actual output parameter in the actual output parameters in the database;
s730, correcting the primary host energy consumption range according to the error between the host output parameter at the preset time and the actual host output parameter to obtain a host energy consumption range at the predicted time;
s800, obtaining the energy consumption range of the water pump according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition specifically comprises the following steps:
s810, acquiring a preliminary water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
s820, acquiring a water pump output parameter in a prediction output parameter of preset time before the prediction time and a water pump actual output parameter in an actual output parameter in the database;
and S830, correcting the preliminary water pump energy consumption range according to the error between the water pump output parameter at the preset time and the actual water pump output parameter to obtain the water pump energy consumption range at the predicted moment.
Specifically, based on step S960, the system uploads the load parameters and the output parameters of the host and the water pump to the database for recording, and then, in the application process of the neural network model (including the neural network load prediction model, the neural network host parameter model and the neural network water pump parameter model), after obtaining the preliminary output results (preliminary predicted air conditioner load, preliminary host energy consumption range and preliminary water pump energy consumption range) based on the external input parameters (the outdoor weather prediction parameters, the host constraint conditions and the water pump constraint conditions), the preliminary output results are further corrected to obtain the final output results (predicted air conditioner load, host energy consumption range and water pump energy consumption range). The correction method comprises the steps of selecting a predicted value and an actual value at any one or more moments before the predicted moment, calculating an error value between the predicted value and the actual value, and correcting according to the error value (selecting an average error value or a weighted average error value at a plurality of moments).
In this embodiment, the system energy-saving group control algorithm is implemented by the cold and heat source machine room energy-saving group control theory, and the neural network model and the group control algorithm are cores, so that the neural network model also needs to correct an output result according to historical prediction data in an application process, thereby improving the accuracy of the prediction result.
In a third embodiment of the present invention, as shown in fig. 6, an energy-saving optimization control system 100 for a cold and heat source machine room includes:
the database module 110 is used for establishing historical data of operating parameters of systems and equipment of a cold and heat source storage machine room in a database;
a model module 120, which takes the load parameters in the database established by the database module 110 as training data to perform offline training and establish a neural network load prediction model;
the model module 120 performs offline training by using the host parameters in the database established by the database module 110 as training data to establish a neural network host parameter model;
the model module 120 is configured to perform offline training by using the water pump parameters in the database established by the database module 110 as training data, and establish a neural network water pump parameter model;
the load prediction module 130 is used for acquiring outdoor weather prediction parameters at the prediction time and acquiring predicted air conditioner load according to the neural network load prediction model established by the model module 120 and the outdoor weather prediction parameters;
the load prediction module 130 specifically includes:
a weather parameter acquiring unit 131 that acquires an outdoor weather prediction parameter at a prediction time;
a load prediction unit 132 for obtaining a preliminary predicted air conditioning load according to the neural network load prediction model and the outdoor weather prediction parameters obtained by the weather parameter obtaining unit 131;
a load parameter acquiring unit 133 that acquires a predicted air-conditioning load and an actual load parameter of a preset time before a predicted time in the database;
a load correction unit 134, configured to correct the preliminary predicted air-conditioning load obtained by the load prediction unit 132 according to the error between the predicted air-conditioning load at the preset time and the actual load parameter obtained by the load parameter obtaining unit 133, so as to obtain a predicted air-conditioning load at a predicted time;
the load transfer module 140 is configured to transfer the predicted air conditioning load obtained by the load prediction module 130 to a host and a water pump to obtain corresponding host constraint conditions and water pump constraint conditions, where the water pump includes a freezing-side water pump and a cooling-side water pump;
a host parameter analysis module 150, configured to obtain a host energy consumption range according to the neural network host parameter model established by the model module 120 and the host constraint condition determined by the load transfer module 140;
the host parameter analysis module 150 specifically includes:
the host energy consumption calculation unit 151 acquires a preliminary host energy consumption range according to the neural network host parameter model and the host constraint condition;
a host parameter obtaining unit 152 that obtains a host output parameter of predicted output parameters at a preset time before a predicted time in the database and a host actual output parameter of actual output parameters;
the host energy consumption correcting unit 153 corrects the preliminary host energy consumption range obtained by the host actual energy consumption calculating unit according to the host output parameter of the preset time obtained by the host parameter obtaining unit 152, so as to obtain a host energy consumption range at a predicted time;
the water pump parameter analysis module 160 obtains a water pump energy consumption range according to the neural network water pump parameter model established by the model module 120, the water pump constraint condition determined by the load transfer module 140, and the host constraint parameter;
the water pump parameter analysis module 160 specifically includes:
the water pump energy consumption calculation unit 161 is used for obtaining a preliminary water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint conditions and the host constraint conditions;
a water pump parameter obtaining unit 162 that obtains a water pump output parameter of the predicted output parameters and a water pump actual output parameter of the actual output parameters at a preset time before the predicted time in the database;
the water pump energy consumption correcting unit 163 corrects the preliminary water pump energy consumption range obtained by the water pump energy consumption calculating unit 161 according to the error between the water pump output parameter at the preset time and the actual water pump output parameter obtained by the water pump parameter obtaining unit 162, so as to obtain a water pump energy consumption range at a predicted time;
the output parameter analysis module 170 determines the predicted output parameters of the host and the water pump under the system energy-saving group control strategy through multiple iterative calculations according to the host energy consumption range determined by the host parameter analysis module 150 and the water pump energy consumption range determined by the water pump parameter analysis module 160;
a parameter information acquisition module 180 for acquiring parameter information of the predicted time through a sensor;
the actual parameter analysis module 185 calculates actual air conditioner load and actual output parameters of the host and the water pump according to the parameter information acquired by the parameter information acquisition module 180;
the processing module 190 uploads the predicted air conditioning load obtained by the load prediction module 130, the predicted output parameter obtained by the output parameter analysis module 170, the obtained actual air conditioning load and the actual output parameter to the database established by the database module 110.
The specific operation modes of the modules in this embodiment have been described in detail in the corresponding method embodiments, and thus are not described in detail again.
A fourth embodiment of the present invention, as shown in fig. 7, is a method and a system for energy-saving optimization control of a cold and heat source machine room, including: the system comprises a database, a neural network load prediction model, a neural network host parameter model, a neural network water pump parameter model and a system energy-saving group control algorithm, wherein the historical database is a foundation, and the neural network model and the group control algorithm are cores.
The actual characteristics of equipment and a system group control energy-saving theory are deeply combined, and the cold and heat source machine room is operated under the required load in a most energy-saving mode under the condition of ensuring the safe operation of the system and the equipment. Firstly, a neural network model for load prediction and neural network parameter models of a host and a water pump are established, and off-line training is performed by using a large amount of historical data to obtain the characteristics of building load and equipment accuracy. And secondly, calculating the load of the building in the next day in real time through outdoor meteorological environment parameters (such as weather forecast). And then, distributing the host and the water pump through an energy-saving group control strategy, controlling the host and the water pump within a safe constraint condition, and calculating the output of the host and the water pump in real time to enable the refrigerating capacity or the heating capacity output by the equipment to meet the load requirement. And finally, outputting the data to a controller of the equipment, executing the equipment according to the calculated result, and recording the data.
The high COP of the host does not imply a certain power saving. No matter how the COP of the host computer changes, under the same working condition, the power of the host computer increases with the increase of the refrigerating capacity, as shown in fig. 8, if the refrigerating capacity of the host computer exceeds the required load, part of the refrigerating capacity is wasted, even if the COP is high, the energy consumption may be higher than that under the required load, and the problem that the refrigerating capacity is higher than the required load can be caused by intentionally operating the host computer in an area with higher COP, so that the energy waste is caused.
Similarly, the water pump is also in a similar situation, as shown in fig. 9, under the same other conditions, the energy consumption of the water pump also increases with the increase of the flow rate, and an appropriate flow rate should be selected, so that the flow rate does not need to be increased intentionally to operate the water pump in a high-efficiency area.
According to the reverse carnot cycle principle, the lower the cooling water temperature, the higher the chilled water temperature, the lower the power of the main machine under the same load, and the higher the COP of the main machine, fig. 10 shows the relationship between the cooling water temperature and the main machine and the cooling water pump group, when other parameters are not changed, the power of the main machine rises with the rise of the cooling water temperature, the power of the cooling water pump group decreases with the height of the cooling water temperature, and an optimal cooling water temperature exists to enable the sum of the total power of the main machine and the water pump group to be the lowest. The same applies to the frozen side as shown in fig. 11.
The final output control object of the energy-saving group control method provided by the invention comprises the following steps: predicted load value, number of hosts, set temperature, number of cooling water, cooling pump frequency, number of chilled water pumps, chilled water pump frequency, bypass valve opening, and the like.
It should be noted that the above embodiments can be freely combined as necessary. The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Claims (10)
1. An energy-saving optimization control method for a cold and heat source machine room is characterized by comprising the following steps:
establishing historical data of operating parameters of systems and equipment of a cold and heat source storage machine room in a database;
taking the load parameters in the database as training data to perform off-line training, and establishing a neural network load prediction model;
taking the host parameters in the database as training data to perform off-line training, and establishing a neural network host parameter model;
taking the water pump parameters in the database as training data to perform off-line training, and establishing a neural network water pump parameter model; the neural network water pump parameter model is used for summarizing the flow and power consumption conditions of the water pumps under different numbers and different frequencies;
acquiring outdoor weather prediction parameters at the prediction time, and acquiring predicted air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters;
transmitting the predicted air conditioner load to a host and a water pump to respectively obtain corresponding host constraint conditions and water pump constraint conditions, wherein the water pump comprises a freezing side water pump and a cooling side water pump;
obtaining a host energy consumption range according to the neural network host parameter model and the host constraint condition;
obtaining a water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
and determining the predicted output parameters of the host and the water pump under the energy-saving group control strategy of the system through repeated iterative calculation according to the energy consumption range of the host and the energy consumption range of the water pump.
2. The cold and heat source machine room energy-saving optimization control method according to claim 1, wherein after determining the predicted output parameters of the host and the water pump under the system energy-saving group control strategy through multiple iterative calculations according to the host energy consumption range and the water pump energy consumption range, the method further comprises:
acquiring parameter information of a predicted moment through a sensor;
calculating actual air conditioner load and actual output parameters of the host and the water pump according to the parameter information;
and uploading the predicted air conditioner load, the predicted output parameters, the actual air conditioner load and the actual output parameters to the database.
3. The cold and heat source machine room energy-saving optimization control method according to claim 2, wherein obtaining outdoor weather prediction parameters at a prediction time, and obtaining a predicted air conditioning load according to the neural network load prediction model and the outdoor weather prediction parameters specifically comprises:
acquiring outdoor weather prediction parameters at a prediction moment;
acquiring a preliminary prediction air conditioner load according to the neural network load prediction model and the outdoor weather prediction parameters;
acquiring a predicted air-conditioning load and an actual air-conditioning load of preset time before the predicted time in the database;
and correcting the preliminary predicted air-conditioning load according to the error between the predicted air-conditioning load and the actual air-conditioning load at the preset time to obtain the predicted air-conditioning load at the predicted time.
4. The cold and heat source machine room energy-saving optimization control method according to claim 2, wherein obtaining the host energy consumption range according to the neural network host parameter model and the host constraint condition specifically comprises:
acquiring a primary host energy consumption range according to the neural network host parameter model and the host constraint condition;
obtaining a host output parameter in the predicted output parameters of the preset time before the predicted time and a host actual output parameter in the actual output parameters in the database;
and correcting the primary host energy consumption range according to the error between the host output parameter at the preset time and the actual host output parameter to obtain the host energy consumption range at the predicted moment.
5. The cold and heat source machine room energy-saving optimization control method according to claim 2, wherein obtaining the water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition specifically comprises:
acquiring a preliminary water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
acquiring a water pump output parameter in a prediction output parameter of preset time before the prediction time and a water pump actual output parameter in an actual output parameter in the database;
and correcting the preliminary water pump energy consumption range according to the error between the water pump output parameter at the preset time and the actual water pump output parameter to obtain the water pump energy consumption range at the predicted moment.
6. The utility model provides a cold and heat source computer lab energy-conservation optimizing control system which characterized in that includes:
the database module is used for establishing historical data of operating parameters of systems and equipment of a cold and heat source storage machine room in a database;
the model module is used for performing off-line training by taking the load parameters in the database established by the database module as training data to establish a neural network load prediction model;
the model module is used for performing off-line training by taking the host parameters in the database established by the database module as training data to establish a neural network host parameter model;
the model module is used for performing off-line training by taking the water pump parameters in the database established by the database module as training data to establish a neural network water pump parameter model; the neural network water pump parameter model is used for summarizing the flow and power consumption conditions of the water pumps under different numbers and different frequencies;
the load prediction module is used for acquiring outdoor meteorological prediction parameters at the prediction time and acquiring the predicted air conditioner load according to the neural network load prediction model established by the model module and the outdoor meteorological prediction parameters;
the load transfer module is used for transferring the predicted air conditioner load acquired by the load prediction module to a host and a water pump to respectively acquire corresponding host constraint conditions and water pump constraint conditions, and the water pump comprises a freezing side water pump and a cooling side water pump;
the host parameter analysis module is used for obtaining a host energy consumption range according to the neural network host parameter model established by the model module and the host constraint condition determined by the load transfer module;
the water pump parameter analysis module is used for obtaining a water pump energy consumption range according to the neural network water pump parameter model established by the model module, the water pump constraint condition determined by the load transfer module and the host constraint parameter;
and the output parameter analysis module is used for determining the predicted output parameters of the host and the water pump under the system energy-saving group control strategy through repeated iterative calculation according to the host energy consumption range determined by the host parameter analysis module and the water pump energy consumption range determined by the water pump parameter analysis module.
7. The energy-saving optimization control system for the cold and heat source machine room according to claim 6, further comprising:
the parameter information acquisition module acquires parameter information of the predicted moment through a sensor;
the actual parameter analysis module calculates actual air conditioner load and actual output parameters of the host and the water pump according to the parameter information acquired by the parameter information acquisition module;
and the processing module uploads the predicted air conditioner load obtained by the load prediction module, the predicted output parameter obtained by the output parameter analysis module, the actual air conditioner load obtained by the actual parameter analysis module and the actual output parameter to the database established by the database module.
8. The energy-saving optimization control system for the cold and heat source machine room according to claim 7, wherein the load prediction module specifically comprises:
a weather parameter acquiring unit for acquiring outdoor weather prediction parameters at the prediction time;
the load prediction unit acquires a preliminary predicted air conditioner load according to the neural network load prediction model and the outdoor meteorological prediction parameters acquired by the meteorological parameter acquisition unit;
a load parameter acquiring unit that acquires a predicted air-conditioning load and an actual air-conditioning load of a preset time before a predicted time in the database;
and the load correction unit is used for correcting the preliminary predicted air-conditioning load obtained by the load prediction unit according to the error between the predicted air-conditioning load at the preset time and the actual air-conditioning load obtained by the load parameter acquisition unit to obtain the predicted air-conditioning load at the predicted time.
9. The energy-saving optimization control system for the cold and heat source machine room according to claim 7, wherein the host parameter analysis module specifically comprises:
the host energy consumption calculation unit is used for acquiring a primary host energy consumption range according to the neural network host parameter model and the host constraint condition;
the host parameter acquiring unit is used for acquiring host output parameters in the predicted output parameters of the preset time before the predicted time in the database and host actual output parameters in the actual output parameters;
and the host energy consumption correction unit is used for correcting the primary host energy consumption range obtained by the host energy consumption calculation unit according to the host output parameter of the preset time and the host actual output parameter error obtained by the host parameter obtaining unit to obtain the host energy consumption range at the predicted moment.
10. The energy-saving optimization control system for the cold and heat source machine room according to claim 7, wherein the water pump parameter analysis module specifically comprises:
the water pump energy consumption calculation unit is used for acquiring a preliminary water pump energy consumption range according to the neural network water pump parameter model, the water pump constraint condition and the host constraint condition;
the water pump parameter acquisition unit is used for acquiring a water pump output parameter in the predicted output parameters of the preset time before the predicted time in the database and a water pump actual output parameter in the actual output parameters;
and the water pump energy consumption correction unit is used for correcting the preliminary water pump energy consumption range obtained by the water pump energy consumption calculation unit according to the error between the water pump output parameter of the preset time and the actual water pump output parameter obtained by the water pump parameter obtaining unit to obtain the water pump energy consumption range at the predicted moment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910691078.5A CN110410942B (en) | 2019-07-30 | 2019-07-30 | Energy-saving optimization control method and system for cold and heat source machine room |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910691078.5A CN110410942B (en) | 2019-07-30 | 2019-07-30 | Energy-saving optimization control method and system for cold and heat source machine room |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110410942A CN110410942A (en) | 2019-11-05 |
| CN110410942B true CN110410942B (en) | 2020-12-29 |
Family
ID=68363942
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910691078.5A Active CN110410942B (en) | 2019-07-30 | 2019-07-30 | Energy-saving optimization control method and system for cold and heat source machine room |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110410942B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110779132A (en) * | 2019-11-13 | 2020-02-11 | 垚控科技(上海)有限公司 | Water pump equipment operation control system of air conditioning system based on reinforcement learning |
| CN110878959B (en) * | 2019-11-28 | 2021-04-27 | 济南大学 | Building temperature control method and system based on model predictive control |
| CN111043731A (en) * | 2019-12-31 | 2020-04-21 | 中维通(北京)科技有限公司 | Air conditioning system adjusting method based on weather forecast |
| CN111047117B (en) * | 2019-12-31 | 2021-02-19 | 中维通(北京)科技有限公司 | Differential pressure bypass valve energy-saving optimization method based on terminal load prediction |
| CN111474858B (en) * | 2020-03-31 | 2024-04-19 | 北京天成冠通能源科技有限公司 | Building energy management control system and method |
| CN113757922B (en) * | 2020-06-03 | 2023-04-07 | 中国移动通信集团浙江有限公司 | Deep learning-based air conditioning system energy-saving control method, device and equipment and computer medium |
| CN113776171B (en) * | 2020-06-10 | 2024-02-13 | 中兴通讯股份有限公司 | Refrigeration equipment control method, refrigeration equipment control device, computer equipment and computer readable medium |
| CN111735178A (en) * | 2020-07-15 | 2020-10-02 | 西安木牛能源技术服务有限公司 | Air conditioner energy-saving system based on elite meaning and SVR regression algorithm and optimization method |
| CN112283890A (en) * | 2020-10-26 | 2021-01-29 | 济中节能技术（苏州）有限公司 | Cold and heat quantity control method and device suitable for building heating and ventilation equipment monitoring system |
| CN112665123B (en) * | 2020-12-10 | 2022-04-29 | 珠海格力电器股份有限公司 | Building load prediction method and device and central air conditioner |
| CN112628956B (en) * | 2020-12-24 | 2022-02-15 | 华中科技大学 | Water chilling unit load prediction control method and system based on edge cloud cooperative framework |
| CN113739365A (en) * | 2021-08-31 | 2021-12-03 | 广州汇电云联互联网科技有限公司 | Central air-conditioning cold station group control energy-saving control method, device, equipment and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251291A (en) * | 2008-04-03 | 2008-08-27 | 上海交通大学 | Central air conditioning system global optimization energy-saving control method and device based on model |
| US20160146493A1 (en) * | 2014-11-26 | 2016-05-26 | International Business Machines Corporation | Building thermal control techniques |
| CN108489012A (en) * | 2018-01-30 | 2018-09-04 | 深圳市新环能科技有限公司 | Cold source of air conditioning energy efficiency model control method based on load prediction and constraint |
| CN109253494A (en) * | 2018-09-05 | 2019-01-22 | 东北大学 | A kind of electric heat-storage device control system and method based on heat load prediction |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US10584698B2 (en) * | 2016-04-07 | 2020-03-10 | Schlumberger Technology Corporation | Pump assembly health assessment |
| CN109595747B (en) * | 2018-12-24 | 2020-01-07 | 珠海格力电器股份有限公司 | Energy consumption simulation method and device of air conditioning system |
-
2019
- 2019-07-30 CN CN201910691078.5A patent/CN110410942B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101251291A (en) * | 2008-04-03 | 2008-08-27 | 上海交通大学 | Central air conditioning system global optimization energy-saving control method and device based on model |
| US20160146493A1 (en) * | 2014-11-26 | 2016-05-26 | International Business Machines Corporation | Building thermal control techniques |
| CN108489012A (en) * | 2018-01-30 | 2018-09-04 | 深圳市新环能科技有限公司 | Cold source of air conditioning energy efficiency model control method based on load prediction and constraint |
| CN109253494A (en) * | 2018-09-05 | 2019-01-22 | 东北大学 | A kind of electric heat-storage device control system and method based on heat load prediction |
Non-Patent Citations (2)
| Title |
|---|
| 基于动态反馈的多元参数回归空调负荷预测控制方法;张国华等;《机电工程技术》;20181128;第47卷(第11期);全文 * |
| 基于负荷预测的商场空调冷冻水系统优化运行研究;蔡盼盼;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20151231;C038-497 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110410942A (en) | 2019-11-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110410942B (en) | Energy-saving optimization control method and system for cold and heat source machine room | |
| CN110392515B (en) | Historical data-based cold and heat source machine room energy-saving control method and system | |
| CN110288164B (en) | Predictive control method for building air-conditioning refrigeration station system | |
| CN111256294B (en) | Model prediction-based optimization control method for combined operation of water chilling unit | |
| EP3483517B1 (en) | System and method for predicting load of and controlling subway heating, ventilation, and air conditioning system | |
| CN112503746B (en) | Control method of cold source system of power station house based on machine learning and particle swarm algorithm | |
| CN114383299B (en) | Central air-conditioning system operation strategy optimization method based on big data and dynamic simulation | |
| CN115220351B (en) | Intelligent energy-saving optimization control method for building air conditioning system based on cloud side end | |
| CN114154677A (en) | Air conditioner operation load model construction and prediction method, device, equipment and medium | |
| CN115325682A (en) | Optimization control method and device for performance monitoring of efficient intelligent refrigeration machine room | |
| CN116266253A (en) | Optimization control method, system and computer readable storage medium for air conditioner parameters | |
| CN111664550A (en) | Energy efficiency optimization method and system based on prediction model and optimization algorithm | |
| Nishiguchi et al. | Data-driven optimal control for building energy conservation | |
| CN115081220A (en) | Adjusting method and system for high-energy-efficiency central air-conditioning system | |
| CN115049141A (en) | Control method and control device for refrigeration system and refrigeration system | |
| Deng et al. | Optimal control of chilled water system with ensemble learning and cloud edge terminal implementation | |
| Hussein et al. | Energy saving by reinforcement learning for multi-chillers of HVAC systems | |
| CN111043720B (en) | Low-cost robustness adjustment strategy making method of refrigeration system under load uncertainty | |
| CN114206071A (en) | Data center cooling system based on hyperbolic cooling tower | |
| CN103528294B (en) | A kind of efficiency processing method of refrigeration system and system | |
| CN112432269A (en) | Method and system for optimizing set value of pressure difference of refrigerating water pump of refrigerating room | |
| CN117250873B (en) | Data center liquid cooling system parameter optimization method | |
| CN117970986B (en) | Temperature and humidity control method, device and medium of cold and hot system | |
| CN112857078B (en) | Dynamic hydraulic balance adjusting method and system for cooling tower group water system | |
| CN114692493A (en) | Energy-saving method and system for building refrigeration system and electronic equipment |
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 |