CN112329338A - Cold source system control method and device based on fish swarm optimization BP neural network - Google Patents

Cold source system control method and device based on fish swarm optimization BP neural network Download PDF

Info

Publication number
CN112329338A
CN112329338A CN202011158663.8A CN202011158663A CN112329338A CN 112329338 A CN112329338 A CN 112329338A CN 202011158663 A CN202011158663 A CN 202011158663A CN 112329338 A CN112329338 A CN 112329338A
Authority
CN
China
Prior art keywords
cold source
neural network
source system
data
fish swarm
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
CN202011158663.8A
Other languages
Chinese (zh)
Other versions
CN112329338B (en
Inventor
刘天伟
李程贵
谢昆
王智慧
马宇晴
侯瑞强
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.)
China Mobile Communications Group Co Ltd
China Mobile Group Inner Mongolia Co Ltd
Original Assignee
China Mobile Communications Group Co Ltd
China Mobile Group Inner Mongolia Co Ltd
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 China Mobile Communications Group Co Ltd, China Mobile Group Inner Mongolia Co Ltd filed Critical China Mobile Communications Group Co Ltd
Priority to CN202011158663.8A priority Critical patent/CN112329338B/en
Publication of CN112329338A publication Critical patent/CN112329338A/en
Application granted granted Critical
Publication of CN112329338B publication Critical patent/CN112329338B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/08—Thermal analysis or thermal optimisation

Abstract

The embodiment of the disclosure provides a cold source system control method and device based on a fish swarm optimization BP neural network, wherein the method comprises the steps of establishing a cold source system control model according to the relation between various types of regulation influence factors and cold source output parameters in a cold source system; collecting data of the adjusting influence factors and cold source output parameters to form a data set; optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set; constructing a neural network, inputting the optimal weight value combination set into the neural network for training so as to be used for automatic control of a cold source system; according to the method, the fish swarm algorithm and the neural network are combined, the collected data can be quickly optimized and trained, the actual output of the cold source system is infinitely close to the expected value of the system, and the intelligent control of the ideal temperature of the system can be more effectively realized.

Description

Cold source system control method and device based on fish swarm optimization BP neural network
Technical Field
The disclosure belongs to the field of data center cold source systems and artificial intelligence, and particularly relates to a cold source system control method and device based on a fish swarm optimization BP neural network.
Background
The use method of the cold source system of the existing data center is to manually control the opening and closing of the cooling source system equipment, and judge the opening quantity and the type of the cold source equipment according to the external temperature and the tail end load capacity through the experience of a maintainer, so that the maintainer can manually open or close the cold source system equipment; meanwhile, the selection of the operation scheme of the cold source system equipment also depends on the experience of maintenance personnel.
At present, no intelligent control is used for the on-off scheme of cold source system equipment; the manual operation mechanism is limited by experience, working state, capability and the like of maintenance personnel, so that the situations that a water chilling unit and a plate heat exchanger are not properly used and the supply of a cold source at the tail end is unstable due to manual operation errors occur occasionally, and then high temperature and downtime of equipment with loads at the tail end are caused; this constitutes the hidden danger to the steady operation of data center's business, and the urgent need is now optimized the use method of cold source system to realize cold source system reasonable steady operation.
Therefore, the prior art is easy to have the condition of high-temperature alarm of the terminal equipment, and can not meet the requirement of the stability of cold source supply.
In view of this, the present disclosure is set forth.
Disclosure of Invention
The embodiment of the disclosure provides a cold source system control method, a cold source system control device, cold source system equipment and a computer storage medium based on a fish swarm optimization BP neural network, and the purpose of improving the stability of a cold source can be achieved through intelligent control over each piece of equipment of the cold source system.
In a first aspect, an embodiment of the present disclosure provides a cold source system control method based on a fish swarm optimization BP neural network, where the method includes:
establishing a cold source system control model according to the relationship between the multi-class regulation influence factors and the cold source output parameters in the cold source system;
collecting data of the adjusting influence factors and cold source output parameters to form a data set;
optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set;
and constructing a neural network, and inputting the optimal weight value combination set into the neural network for training so as to be used for automatic control of a cold source system.
In some embodiments, the optimizing the data set by the fish swarm algorithm to obtain an optimal weight combination set in the data set includes
Setting basic parameters of a fish school algorithm according to the data volume of the data set, wherein the basic parameters at least comprise fish school rules, step length, visual field and crowding degree factors;
and calculating the output parameters of each cold source group and the data of various corresponding adjusting influence factors in the data set by a fish swarm algorithm to obtain the optimal weight of various data, and forming an optimal weight combination set for inputting into a neural network for training.
In some embodiments, the fish swarm algorithm includes foraging behavior, herding behavior, tailgating behavior, and/or stochastic behavior.
In some embodiments, the constructing a neural network, inputting the optimal weight value combination set into the neural network for training, includes:
setting the number of layers of a neural network, output neurons and input neurons according to the relationship between the multiple types of regulation influence factors and the cold source output parameters in the cold source system, and constructing the neural network of the cold source system;
setting standard cold source output parameters of a cold source system, constructing a relation between real cold source output parameters and the standard output parameters according to the minimum mean square error, and deducing a weight value adjusting formula according to the relation;
and inputting the data in the optimal weight combination set into a neural network, training the neural network through a weight adjustment formula until the error between the standard cold source output parameter in the neural network and the cold source output parameter in the optimal weight combination set is within a preset range, and finishing training.
In some embodiments, the inputting the optimal weight value combination set into the neural network for training for automatic control of the heat sink system includes
Inputting a part of adjusting influence factor values in actual operation into the trained neural network, so that the neural network outputs at least one cold source output parameter;
and the cold source system control model reversely calculates the values of the rest adjusting influence factors according to the cold source output parameters and the values of the partial adjusting influence factors in the actual operation, and adjusts and controls the cold source system.
In some embodiments, the cold source system is a water cooling system, the multiple types of adjustment influence factors include an outdoor temperature, a terminal load amount, and a number of cold source devices turned on, and the cold source output parameter is an outlet water temperature of the chilled water.
In some embodiments, the method further comprises establishing a cold source system control model based on the relationship between the plurality of types of modulation influencing factors and the output parameters of the cold source in the cold source system, including
Presetting a linkage condition of a water chilling unit and heat exchange equipment with temperature as a limit value, so that the heat exchange equipment works independently when the linkage condition is lower than the linkage condition; when the linkage condition is met, the water chilling unit and the heat exchange equipment work in a linkage manner; and when the linkage condition is exceeded, the refrigerating unit works independently.
On the other hand, the embodiment of the disclosure also provides a cold source system control device based on the fish swarm optimization BP neural network, and the device comprises
The model construction module is used for establishing a cold source system control model according to the relation between the multi-class regulation influence factors and the cold source output parameters in the cold source system;
the data acquisition module is used for acquiring the data of the adjusting influence factors and the cold source output parameters to form a data set or automatically control the system;
the data optimization module is used for optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set;
and the data learning module is used for constructing a neural network and inputting the optimal weight value combination set into the neural network for training so as to automatically control the cold source system.
In another aspect, an embodiment of the present disclosure further provides a cold source system control device based on a fish swarm optimization BP neural network, where the device includes a processor and a memory storing computer program instructions; the processor reads and executes the computer program instructions to realize the cold source system control method based on the fish swarm optimization BP neural network as described in any embodiment above.
In still another aspect, an embodiment of the present disclosure further provides a computer storage medium, where computer program instructions are stored on the computer storage medium, and when the computer program instructions are executed by a processor, the method for controlling a cold source system based on a fish swarm optimization BP neural network according to any embodiment of the present disclosure is implemented.
The cold source system control method, the device, the equipment and the computer storage medium based on the fish swarm optimization BP neural network can carry out rapid optimization and training on the acquired data by combining the fish swarm algorithm and the neural network, so that the actual output of the cold source system is infinitely close to the expected value of the system, and the intelligent control of the ideal temperature of the system can be more effectively realized.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings needed to be used in the embodiments of the present disclosure will be briefly described below, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flowchart of a cold source system control method based on a fish swarm optimization BP neural network according to an embodiment of the present disclosure;
FIG. 2 is a schematic diagram illustrating the operation of a data center cold source system according to an embodiment of the present disclosure;
fig. 3 is a control logic of a cold source control system constructed in the cold source system control method based on the fish swarm optimization BP neural network provided in the embodiment of the present disclosure;
fig. 4 is a schematic diagram of a selected neural network structure in the cold source system control method based on the fish swarm optimization BP neural network provided in the embodiment of the present disclosure;
fig. 5 is a line graph of a result of performing algorithm verification in the cold source system control method based on the fish swarm optimization BP neural network provided in the embodiment of the present disclosure, where 5a is a result of performing only neural network training; 5b is the result of neural network training after optimization of fish swarm algorithm;
fig. 6 is a flowchart illustrating a cold source system control method based on a fish swarm optimization BP neural network according to an embodiment of the present disclosure;
fig. 7 is a schematic structural diagram of a cold source system control device based on a fish swarm optimization BP neural network according to an embodiment of the present disclosure;
fig. 8 is a schematic structural diagram of a cold source system control device based on a fish swarm optimization BP neural network according to an embodiment of the present disclosure.
Detailed Description
Features and exemplary embodiments of various aspects of the present disclosure will be described in detail below, and in order to make objects, technical solutions and advantages of the present disclosure more apparent, the present disclosure will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are intended to be illustrative only and are not intended to be limiting of the disclosure. It will be apparent to one skilled in the art that the present disclosure may be practiced without some of these specific details. The following description of the embodiments is merely intended to provide a better understanding of the present disclosure by illustrating examples of the present disclosure.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The embodiment of the disclosure provides a cold source system control method, a cold source system control device and a computer storage medium based on a fish swarm optimization BP neural network, wherein the cold source system of a data center is controlled, different characteristics are input, the neural network operation is performed, and finally, the intelligent control of the cold source system, especially a water cooling system is realized, so that the intelligent control of the opening number of a water chilling unit and a plate heat exchanger, the matching use of the water chilling unit and the plate heat exchanger and other conditions are realized, and the reasonable and stable operation of the cold source system is further realized.
First, a cold source system control method based on a fish swarm optimization BP neural network provided by the embodiment of the present disclosure is introduced below.
Fig. 1 shows a flow chart of a cold source system control method based on a fish swarm optimization BP neural network according to an embodiment of the present disclosure. As shown in fig. 1, the method may include the steps of:
s001, establishing a cold source system control model according to the relation between the multi-class regulation influence factors and the cold source output parameters in the cold source system;
s002, collecting data of the adjusting influence factors and cold source output parameters to form a data set;
s003, optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set;
and S004, constructing a neural network, and inputting the optimal weight value combination set into the neural network for training so as to be used for automatic control of a cold source system.
The cold source system in the method disclosed by the invention is a data center cold source system, and the data center cold source system is divided into an air cooling system and a water cooling system, wherein the water cooling system has high efficiency and mainly comprises a water chilling unit, a water pump, a plate heat exchanger (called as 'plate exchanger' for short), a cooling tower and the like, as shown in fig. 2, when the plate exchanger is linked with the water chilling unit, the chilled water pump 100 works, water subjected to heat exchange by the plate exchanger 101 enters a machine room through a water separator 102, after heat exchange is carried out between a water cooling system air conditioner 103 serving as a terminal load in the machine room and the environment of the machine room, the heat exchanged water flows back to a water collector 104, and water in the water collector 104 enters the plate exchanger 101 again after being treated by; when only the plate exchanger 101 is opened, the water chilling unit 105 serves as a pipeline 107 but does not work; only when the water chiller 105 is operating, the plate exchanger 102 is functioning as a conduit but is not operating.
At present, most of large data centers adopt a water cooling system, so the cold source system in the embodiment is described as a water cooling system.
For a large data center, most of the air conditioners of the tail end water cooling systems of the large data center enter a machine room through pipelines, so that heat exchange is carried out between the air conditioners and the environment of the machine room, heat in the machine room is taken away, and the loads such as servers and the like guarantee a proper temperature, so that the large data center is not prone to breakdown due to high temperature. Therefore, when the data center adopts a water cooling system, the outlet water temperature of the chilled water is particularly important for the machine room environment, so that the outlet water temperature of the chilled water can be determined as an output variable by the cold source output parameters in the system, and a control model of the water cooling system is constructed together with other input variables.
S001, establishing a cold source system control model according to the relation between the multi-class regulation influence factors and the cold source output parameters in the cold source system, including
S101, determining the relationship between multiple types of adjusting influence factors and cold source output parameters in a cold source system;
the relationship of the various adjusting influence factors can be parameters influencing the outlet water temperature of the chilled water, and the aim of controlling the outlet water temperature of the chilled water is achieved by adjusting the values of the parameters.
Therefore, the adjusting influence factors comprise one or more of outdoor temperature, tail end load, the starting number of cold source equipment, water flow speed and pipeline pressure; the water flow speed and the pipeline pressure are determined at the beginning of building of the machine room and cannot be changed in actual use, so that the water flow speed and the pipeline pressure are constant values, and the water flow speed and the pipeline pressure can not be considered when a cold source system control model is built; the outdoor temperature, the tail end load and the opening number of cold source equipment all change due to actual conditions and belong to variables, so that the control method of the embodiment needs to optimize the values of the variables, and adjust and control the values of the variables according to the obtained optimal weight, thereby achieving the most reasonable variable control of the temperature of the outlet water of the chilled water; therefore, a cold source system control model is constructed according to the relation between the three regulation influence factors of the outdoor temperature, the tail end load and the opening number of the cold source equipment and the outlet water temperature of the chilled water.
The outdoor temperature changes along with the change of seasons and time in a day, so that the indoor temperature is influenced, the temperature in the machine room also changes, and the temperature of the outlet water of the chilled water also changes along with the change of the outdoor temperature, so that the temperature of the server can be ensured to be within a required range; the load temperature of the machine room except the temperature of the building is determined by the load capacity of the tail end, when the load is large, the temperature generated by the load is increased, and the temperature of the outlet water of the chilled water is required to be correspondingly reduced to ensure that the server can be stably operated; the opening number of cold source equipment (comprising a water chilling unit and a plate exchanger) is the same. In addition, because the northern data center has a natural cold source, namely extremely low outdoor temperature in winter, the cold source equipment generally adopts a mode of linkage operation of a plate heat exchanger and a water chilling unit to maximally utilize the natural cold source so as to achieve the aims of saving energy and reducing PUE.
Therefore, the step S001 further comprises an S102, linkage conditions are set, conditions such as the number of the water chilling units and the plate heat exchangers to be opened and how to match the water chilling units and the plate heat exchangers are configured, intelligent regulation and control are achieved, energy-saving management is conducted on the power consumption of the cold source system on the premise that stable output of the whole system is achieved, and the maximization of the refrigeration efficiency is achieved.
S102, presetting a linkage condition of a water chilling unit and heat exchange equipment with temperature as a limit value, so that the heat exchange equipment works independently when the linkage condition is lower than the limit value; when the linkage condition is met, the water chilling unit and the heat exchange equipment work in a linkage manner; when the linkage condition is exceeded, the refrigerating unit works independently, as shown in fig. 3:
the linkage conditions include:
when the ambient temperature is reduced to below 6 ℃, namely in winter, the heat exchange effect of the plate heat exchanger reaches the maximum; the method comprises the following steps that chilled water entering a machine room tail end water cooling system air conditioner (called tail end load for short, the same below) and cooling water output by a machine room are subjected to heat exchange through a plate heat exchanger, so that the temperature of the outlet water of the chilled water is controlled to be 14 ℃, a cold source is provided for tail end load equipment, and the number of started equipment is calculated according to the cold quantity requirement of the tail end load and the refrigerating capacity of single cold source equipment; under the condition, the cold source equipment is a plate heat exchanger;
when the ambient temperature rises to exceed 6 ℃ and is not higher than 15 ℃, namely in a transition season, the plate heat exchanger and the centrifugal water chilling unit are operated in a linkage mode to cool the chilled water, and at the moment, the cold machine is started to refrigerate under the condition that the refrigerating capacity of the plate heat exchanger cannot be met according to the requirement of the cold quantity loaded at the tail end; calculating the number of the started cold sources according to the end load cold quantity requirement and the refrigerating capacity of a single cold source device; under the condition, cold source equipment, namely a plate heat exchanger and a water chilling unit;
when the ambient temperature rises to more than 15 ℃, namely the condition in summer, the plate heat exchanger has no refrigerating capacity, the water chilling unit is adopted to operate independently to cool the chilled water, and the corresponding number of the water chilling units is started according to the requirement of the load cold quantity at the tail end.
The linkage condition is set, when the subsequent neural network outputs cold source output parameter values (namely the chilled water outlet temperature), the starting condition and the starting quantity of the cold source equipment can be automatically obtained according to the outdoor temperature and the terminal load data, if the plate is opened when the plate works alone, or the plate is opened when the plate works with the water chilling unit in linkage, or the water chilling unit is opened when the water chilling unit works alone, the reasonable energy-saving cold source equipment starting scheme can be rapidly determined, and the intelligent control of a cold source system is achieved.
After the cold source system control model is established, data acquisition is carried out in a targeted manner through the model, and a large number of system related parameters in actual operation are obtained for subsequent neural network training.
S002, collecting data of the adjusting influence factors and cold source output parameters to form a data set;
collecting outdoor temperature Ti(i-1, 2, … …,100), the data is collected quarterly, and the collection frequency is once every half hour;
collecting end load Pi(i-1, 2, … …,100), the data is collected one quarter of a day, and the data collection frequency is once a day because the data center server load changes are small;
cold source equipment turn-on quantity Qq(q 1,2, … …,100) this data is collected quarterly, every half hour, as cold source devices turned on will be affected by outdoor temperature and load.
The regulation influence factor is a factor mainly influencing the outlet water temperature of the chilled water and can be used as an input variable of a BP (back propagation) neural network, so that the data are collected in at least one quarter, and the outlet water temperature T of the chilled water corresponding to the data is collectednData of (1, 2, … …,100), which is collected every half 10 minutes because such data changes frequently. And the tail end water cooling system air conditioner mainly depends on the temperature of the outlet water of the chilled water to refrigerate, so that the tail end water cooling system air conditioner can be used as an output variable of a neural network.
The training mode of neural network self-learning is adopted, partial adjusting influence factors can be analyzed, an optimal cold source output parameter is obtained, the purpose of intelligently controlling a cold source system is achieved, errors of artificial experience judgment are avoided, intelligent and precise system control is achieved, and the method is high in efficiency and good in stability.
In the prior art, the neural network technology is widely applied to different technical fields or different types of neural network structures, for example, in the patent document (publication number: CN104019520B) of "the data driving control method for minimum energy consumption of a refrigeration system based on SPSA", the SPSA algorithm is used to replace the traditional PID control algorithm to control the cooling capacity; or "vehicle air conditioner control system method based on neural network" (CN201310733218.3), the field of using the neural network is in the fields of vehicle and civil; in the two invention application documents, the application field method of the neural network is greatly different from the data center industry, in the existing vehicle air conditioner control system method of the neural network, the application field of the neural network is a vehicle-mounted air conditioner, the aspects of equipment condition, influence factors and the like are different from those of the data center, and the algorithm cannot be used in the data center field; thus, there is currently no application that employs neural networks in data centers.
The intelligent control of the cold source system of the data center adopts the neural network, and the acquired data volume is too large, so that the system is easy to fall into local optimal or optimal impasse when the neural network is trained by directly using a data set, and meanwhile, the convergence speed is low and the solving time is long; therefore, the method further comprises the step S003 of realizing global optimization through fish swarm optimization and artificial fish algorithm and through individual local optimization, importing the optimized data into a neural network, training the neural network, improving training speed and optimizing training effect.
S301, setting basic parameters of a fish swarm algorithm according to the data volume of the data set, wherein the basic parameters at least comprise fish swarm rules, step sizes, visual fields and crowdedness factors.
The principle of the fish school algorithm is to set the current state of an artificial fish as a vector X, wherein X is equal to (X)1,X2,X3…,Xn) Wherein by Xi(i is 1,2,3 …, n), Visual is the Visual field range state of the artificial fish, and X is usedVRepresenting the position of the viewpoint at a certain time, if the state of the position is better than the current state X, the position can be further advanced to the current position direction to reach the state XnextIf the position has a view field status XVAnd if the current state is not better than the current state, the user continues to patrol other positions in the visual field. The surrounding environment is subjected to all-dimensional cognition through tour, and then the optimal value can be found quickly.
In this embodiment, according to the number of collected data, the fish school size N is set to 20, the maximum Step length (Step) S for moving artificial fish is set to 0.3, the maximum Visual field (Visual) V for artificial fish is set to 2.5, and the crowdedness factor δ is set to 3.168; the artificial fish can represent the end load cold quantity requirement, the outdoor temperature, the starting number of refrigeration equipment and the outlet water temperature of chilled water required by a neural network structure corresponding to a cold source system of a data center, and can optimize the four types of data.
S302, calculating each group of cold source output parameters in the data set and the data of various corresponding adjusting influence factors through a fish swarm algorithm to obtain the optimal weight of various data to form an optimal weight combination set;
the food concentration of the current position of the artificial fish is expressed as Y ═ f (x), wherein Y is an objective function value; then, if there is an optimal partner in the current domain (i.e. d < Visual), i.e. there is the data with the highest probability of occurrence among the same kind of data and the food concentration is the greatest, then the partner is set as XbestIf the current position corresponds to the food concentration YbestAccording to Ybest/nf>δYi(nfFor the number of partners in the current domain, δ YiA congestion factor in the current domain), X is indicatedbestWhen the food concentration is high and the food is not crowded, the food concentration is toward XbestThe mathematical expression of (c) is:
Figure BDA0002743629310000101
where the rand () function is a random number between 0 and 1.
The four types of data are optimized according to the method, and the four types of data obtained by optimization are necessarily corresponding to a plurality of groups of data, namely, an optimal chilled water outlet temperature value corresponds to an optimal terminal load cold quantity demand value, an optimal outdoor temperature value and an optimal refrigeration equipment starting quantity value, namely, an optimal weight combination. Carrying out fish school optimization on the whole data set to obtain an optimal weight value combination set; and finishing fish shoal optimization, and outputting the optimal weight combination set to a neural network for training.
If no optimal partner exists in the current field (namely d is less than Visual), variable parameters such as the terminal load cold quantity demand, the outdoor temperature, the starting number of the refrigeration equipment, the chilled water outlet water temperature and the like are respectively set as the individual state vectors of the artificial fish, and the artificial fish is respectively subjected to foraging, clustering, rear-end collision and random behaviors until the state vectors reach XbestAnd then the optimal weight combination is output to the BP neural network to train the BP neural network.
S401, setting the number of layers of a neural network, output neurons and input neurons according to the relation between multiple types of regulation influence factors and cold source output parameters in a cold source system, and constructing the neural network of the cold source system;
the neural network mainly comprises an input layer, a hidden layer and an output layer, and in this embodiment, a BP neural network (a Back Propagation, a multi-layer feedforward network trained according to an error inverse Propagation algorithm) is selected as the structure shown in fig. 4.
The BP neural network of the cold source system is designed into a 3-layer neural network, 3 input neurons and 1 output neuron, and as the input vectors are numbers between [ -1,1] and the distribution of the output vectors exceeds [ -1,1], the tan sig function is selected for hidden layer neuron transfer and the purelin function is selected for input layer neurons. According to the hidden layer neuron formula:
Figure BDA0002743629310000111
in the formula: n is1Is the number of neurons in the hidden layer, nIs the number of neurons in the input layer, s is the number of neurons in the output layer, c is [1,10 ]]Constant in between.
Therefore, the number of the hidden layer neurons is preliminarily judged to be 8; through experiments, when Matlab is programmed, a cyclic function is used, 7-11 is tried to be used as the number of neurons in the hidden layer, 0.001 is respectively used as error precision, smooth convergence can be achieved, but the convergence is fastest at 8, and therefore the number of the neurons in the hidden layer is determined to be 8.
By adopting the fish school scale, the state X of the artificial fish is set as a D-dimensional vector, D is the sum of the weight of the neural network and the threshold, and in the 3-layer neural network in the embodiment, D is (s + n) n1+n1+ s. And after the D-dimension vector is determined, inputting the optimal weight combination set trained by using a fish swarm algorithm into the constructed BP neural network, training the neural network, and finishing the operation between the input and the output of the BP neural network.
S402, standard cold source output parameters of a cold source system are set, a relation between real cold source output parameters and the standard output parameters is constructed according to the minimum mean square error, and a weight value adjusting formula is deduced according to the relation, wherein the weight value adjusting formula comprises the following steps:
if a represents the real output parameter of the output layer and t represents the standard output parameter of the network, the minimum mean square error can be expressed as:
F(x)＝E(e2)＝E((t-a)2)
the weight value adjustment formula is as follows:
Figure BDA0002743629310000112
Figure BDA0002743629310000113
where w is the connection weight between neurons, m is the layer in which it is located,
Figure BDA0002743629310000114
is the weight that the ith neuron is connected to the jth neuron,
Figure BDA0002743629310000115
is the bias of the ith neuron in m layers, alpha is the iteration step size, and k is the iteration step number.
The specific calculation process of the adjustment weight is as follows:
Figure BDA0002743629310000116
Figure BDA0002743629310000117
Figure BDA0002743629310000121
used to refer to the equation after the equal sign "═ for substitution in the equations below.
The following two formulas are obtained:
Figure BDA0002743629310000122
obtaining a gradient descent method adjustment formula of the weight w and the bias b:
Figure BDA0002743629310000123
Figure BDA0002743629310000124
wherein the content of the first and second substances,
Figure BDA0002743629310000125
and S403, inputting the data in the optimal weight combination set into a neural network, training the neural network through the weight adjustment formula until the error between the standard cold source output parameter in the neural network and the cold source output parameter in the optimal weight combination set is within a preset range, and finishing training.
In the neural network training process, comparing the expected value (namely the standard cold source output parameter) of the temperature of the outlet water of the chilled water with the real output value, and if the expected value is close to the real output value, finishing learning; otherwise, calculating error values of each layer in a reverse mode, adjusting weight values w and b of each layer, and carrying out BP neural network training again.
S404, carrying out neural network control on a cold source system
With reference to fig. 6, after training is finished, the cold source system control model is combined with the BP neural network optimized by the fish swarm and linked with the cold source system control model. Outdoor temperature and tail end load capacity are used as chilled water temperature control instructions U1 and U2, the chilled water temperature control instructions are input to a neural network, and a chilled water outlet temperature value is output by the system, so that the starting number of cold source equipment is reversely adjusted, and the actual output value of the chilled water is close to the expected value of the chilled water.
Therefore, the embodiment also verifies the scheme of the disclosure:
1) test environment
In an IDC (Internet Data Center ) cold source system, 3 coolers and plate heat exchangers are used for refrigerating, wherein the return water temperature of chilled water (the temperature is determined by the working conditions of the plate heat exchanger and the water chilling unit in a water cooling system), the tail end load and the outdoor temperature act together to control the outlet water temperature of the chilled water, according to the standard of design Specification for electronic information System (GB 50174-2008), the supply air temperature of a tail end air conditioner is required to be effectively controlled within the range of 18-28 ℃, and the outlet water temperature of the chilled water is required to be controlled within the range of 13-17 ℃.
2) Algorithm checking
After optimizing the data of on-the-spot new acquisition and neural network training, the condition of opening when the adoption cold source equipment kind that reachs and quantity all are controlled with the manual work is similar, example: when the load at the tail end is 10000KW, the outdoor temperature is minus 10 degrees and the temperature of the outlet water of the chilled water is 15 degrees, the system adopts the 2-platen heat exchanger to refrigerate the tail end, the operation result basically conforms to the field test condition, the running model of the cold source system is met, and the normal running of the load at the tail end can be ensured.
Comparing the BP neural network with the BP neural network optimized by the fish swarm algorithm, the neural network training result shown in fig. 5a and the fish swarm optimization neural network training result shown in fig. 5b show that:
as can be seen from the network training diagram: similarly, 0.0001 is an error value, the ordinary BP neural training can reach the target after 168 times of training, and the BP neural network optimized by the fish swarm algorithm can reach the target only by 88 times of training. The latter has a faster convergence characteristic and can find an optimal point in a short time.
Therefore, in the cold source system control method based on the fish swarm optimization BP neural network, the neural network is established through characteristics (including the type and quantity measuring standard of a refrigerating unit and the like) such as outdoor temperature, tail end load capacity, influence relation between the starting quantity of cold source equipment and the outlet water temperature of chilled water, so that artificial intelligent control of a cold source system is realized, high reliability of the system is realized, a stable cold source is provided for tail end load, and the stability of the tail end load environment is improved;
in the training process of using the neural network, the fish swarm algorithm is combined with the BP neural network, so that the fast convergence optimization of the weight of the data is promoted, the operation time of the BP neural network is shortened, and the accuracy of the operation result is promoted.
In a second aspect, the invention also provides a cold source system control device based on the fish swarm optimization BP neural network, which can implement the method, as shown in fig. 7, the device comprises
The model construction module is used for establishing a cold source system control model according to the relation between the multi-class regulation influence factors and the cold source output parameters in the cold source system;
the data acquisition module is used for acquiring the data of the adjusting influence factors and the cold source output parameters to form a data set or automatically control the system;
the data optimization module is used for optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set;
and the data learning module is used for constructing a neural network and inputting the optimal weight value combination set into the neural network for training so as to automatically control the cold source system.
Each module in the apparatus shown in fig. 7 has a function of implementing each step in fig. 1, and can achieve the corresponding technical effect, and for brevity, no further description is given here.
In another aspect, an embodiment of the present disclosure provides a cold source system control device based on a fish swarm optimization BP neural network, as shown in fig. 8, which is a schematic structural diagram of the device.
The cold source system control device based on the fish swarm optimization BP neural network may comprise a processor 301 and a memory 302 storing computer program instructions.
In particular, the processor 301 may include a Central Processing Unit (CPU), an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits configured to implement the embodiments of the present disclosure.
Memory 302 may include mass storage for data or instructions. By way of example, and not limitation, memory 302 may include a Hard Disk Drive (HDD), floppy Disk Drive, flash memory, optical Disk, magneto-optical Disk, tape, or Universal Serial Bus (USB) Drive or a combination of two or more of these. In one example, memory 302 can include removable or non-removable (or fixed) media, or memory 302 is non-volatile solid-state memory. The memory 302 may be internal or external to the integrated gateway disaster recovery device.
The processor 301 reads and executes the computer program instructions stored in the memory 302 to implement the methods/steps S001 to S004 in the embodiment shown in fig. yy, and achieve the corresponding technical effects achieved by the embodiment shown in fig. 1 executing the methods/steps thereof, which are not described herein again for brevity.
In one example, the cold source system control device based on the fish swarm optimization BP neural network may further include a communication interface 303 and a bus 310. As shown in fig. 3, the processor 301, the memory 302, and the communication interface 303 are connected via a bus 310 to complete communication therebetween.
The communication interface 303 is mainly used for implementing communication between modules, apparatuses, units and/or devices in the embodiments of the present disclosure.
Bus 310 includes hardware, software, or both to couple the components of the online data traffic billing device to each other. By way of example, and not limitation, a Bus may include an Accelerated Graphics Port (AGP) or other Graphics Bus, an Enhanced Industry Standard Architecture (EISA) Bus, a Front-Side Bus (Front Side Bus, FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) Bus, an infiniband interconnect, a Low Pin Count (LPC) Bus, a memory Bus, a Micro Channel Architecture (MCA) Bus, a Peripheral Component Interconnect (PCI) Bus, a PCI-Express (PCI-X) Bus, a Serial Advanced Technology Attachment (SATA) Bus, a video electronics standards association local (VLB) Bus, or other suitable Bus or a combination of two or more of these. Bus 310 may include one or more buses, where appropriate. Although this disclosed embodiment describes and illustrates a particular bus, this disclosure contemplates any suitable bus or interconnect.
In addition, in combination with the cold source system control method based on the fish swarm optimization BP neural network in the above embodiments, the embodiments of the present disclosure may provide a computer storage medium to implement. The computer storage medium having computer program instructions stored thereon; when being executed by a processor, the computer program instructions implement any one of the above-mentioned cold source system control methods based on the fish swarm optimization BP neural network.
It is to be understood that this disclosure is not limited to the particular configurations and processes described above and shown in the drawings. A detailed description of known methods is omitted herein for the sake of brevity. In the above embodiments, several specific steps are described and shown as examples. However, the method processes of the present disclosure are not limited to the specific steps described and illustrated, and those skilled in the art may make various changes, modifications, and additions or change the order between the steps after comprehending the spirit of the present disclosure.
The functional blocks shown in the above-described structural block diagrams may be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, it may be, for example, an electronic Circuit, an Application Specific Integrated Circuit (ASIC), suitable firmware, plug-in, function card, or the like. When implemented in software, the elements of the present disclosure are the programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine-readable medium or transmitted by a data signal carried in a carrier wave over a transmission medium or a communication link. A "machine-readable medium" may include any medium that can store or transfer information. Examples of a machine-readable medium include electronic circuits, semiconductor memory devices, ROM, flash memory, Erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, Radio Frequency (RF) links, and so forth. The code segments may be downloaded via computer networks such as the internet, intranet, etc.
It should also be noted that the exemplary embodiments mentioned in this disclosure describe some methods or systems based on a series of steps or devices. However, the present disclosure is not limited to the order of the above-described steps, that is, the steps may be performed in the order mentioned in the embodiments, may be performed in an order different from the order in the embodiments, or may be performed several steps at the same time.
Aspects of the present disclosure are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions/acts specified in the flowchart and/or block diagram block or blocks. Such a processor may be, but is not limited to, a general purpose processor, a special purpose processor, an application specific processor, or a field programmable logic circuit. It will also be understood that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware for performing the specified functions or acts, or combinations of special purpose hardware and computer instructions.
As described above, only the specific embodiments of the present disclosure are provided, and it can be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the system, the module and the unit described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again. It should be understood that the scope of the present disclosure is not limited thereto, and any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope of the present disclosure, and these modifications or substitutions should be covered within the scope of the present disclosure.

Claims (10)

1. A cold source system control method based on a fish swarm optimization BP neural network is characterized by comprising the following steps:
establishing a cold source system control model according to the relationship between the multi-class regulation influence factors and the cold source output parameters in the cold source system;
collecting data of the adjusting influence factors and cold source output parameters to form a data set;
optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set;
and constructing a neural network, and inputting the optimal weight value combination set into the neural network for training so as to be used for automatic control of a cold source system.
2. The cold source system control method based on the fish swarm optimization BP neural network as claimed in claim 1, wherein the optimization of the data set through the fish swarm algorithm to obtain the optimal weight combination set in the data set comprises
Setting basic parameters of a fish school algorithm according to the data volume of the data set, wherein the basic parameters at least comprise fish school rules, step length, visual field and crowding degree factors;
and calculating the output parameters of each cold source group and the data of various corresponding adjusting influence factors in the data set by a fish swarm algorithm to obtain the optimal weight of various data, and forming an optimal weight combination set for inputting into a neural network for training.
3. The cold source system control method based on the fish swarm optimization BP neural network according to claim 2, wherein the fish swarm algorithm comprises foraging behavior, clustering behavior, tailgating behavior and/or random behavior.
4. The cold source system control method based on the fish swarm optimization BP neural network as claimed in claim 2, wherein the constructing the neural network, inputting the optimal weight combination set into the neural network for training, comprises:
setting the number of layers of a neural network, output neurons and input neurons according to the relationship between the multiple types of regulation influence factors and the cold source output parameters in the cold source system, and constructing the neural network of the cold source system;
setting standard cold source output parameters of a cold source system, constructing a relation between real cold source output parameters and the standard output parameters according to the minimum mean square error, and deducing a weight value adjusting formula according to the relation;
and inputting the data in the optimal weight combination set into a neural network, training the neural network through a weight adjustment formula until the error between the standard cold source output parameter in the neural network and the cold source output parameter in the optimal weight combination set is within a preset range, and finishing training.
5. The method as claimed in claim 4, wherein the method for controlling a cold source system based on a fish school optimization BP neural network comprises inputting the optimal weight combination set into the neural network for training, including
Inputting a part of adjusting influence factor values in actual operation into the trained neural network, so that the neural network outputs at least one cold source output parameter;
and the cold source system control model reversely calculates the values of the rest adjusting influence factors according to the cold source output parameters and the values of the partial adjusting influence factors in the actual operation, and adjusts and controls the cold source system.
6. The method as claimed in any one of claims 1 to 5, wherein the cold source system is a water cooling system, the plurality of types of adjustment influencing factors include outdoor temperature, terminal load amount and the number of cold source devices turned on, and the output parameter of the cold source is the outlet temperature of the chilled water.
7. The method as claimed in claim 6, wherein the method for controlling the cooling source system based on the fish swarm optimization BP neural network comprises establishing a cooling source system control model according to the relationship between the multiple types of adjusting influence factors and the output parameters of the cooling source in the cooling source system, including
Presetting a linkage condition of a water chilling unit and heat exchange equipment with temperature as a limit value, so that the heat exchange equipment works independently when the linkage condition is lower than the linkage condition; when the linkage condition is met, the water chilling unit and the heat exchange equipment work in a linkage manner; and when the linkage condition is exceeded, the refrigerating unit works independently.
8. A cold source system control device based on a fish swarm optimization BP neural network is characterized by comprising
The model construction module is used for establishing a cold source system control model according to the relation between the multi-class regulation influence factors and the cold source output parameters in the cold source system;
the data acquisition module is used for acquiring the data of the adjusting influence factors and the cold source output parameters to form a data set or automatically control the system;
the data optimization module is used for optimizing the data set through a fish swarm algorithm to obtain an optimal weight combination set in the data set;
and the data learning module is used for constructing a neural network and inputting the optimal weight value combination set into the neural network for training so as to automatically control the cold source system.
9. A cold source system control device based on a fish swarm optimization (BP) neural network, which is characterized by comprising a processor and a memory, wherein computer program instructions are stored in the memory; the processor reads and executes the computer program instructions to implement the method for controlling the cold source system based on the fish swarm optimization BP neural network as claimed in any one of claims 1 to 7.
10. A computer storage medium, wherein the computer storage medium stores thereon computer program instructions, and when the computer program instructions are executed by a processor, the computer program instructions implement the cold source system control method based on the fish swarm optimization BP neural network according to any one of claims 1 to 7.
CN202011158663.8A 2020-10-26 2020-10-26 Cold source system control method and device based on fish swarm optimization BP neural network Active CN112329338B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011158663.8A CN112329338B (en) 2020-10-26 2020-10-26 Cold source system control method and device based on fish swarm optimization BP neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011158663.8A CN112329338B (en) 2020-10-26 2020-10-26 Cold source system control method and device based on fish swarm optimization BP neural network

Publications (2)

Publication Number Publication Date
CN112329338A true CN112329338A (en) 2021-02-05
CN112329338B CN112329338B (en) 2023-07-28

Family

ID=74311920

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011158663.8A Active CN112329338B (en) 2020-10-26 2020-10-26 Cold source system control method and device based on fish swarm optimization BP neural network

Country Status (1)

Country Link
CN (1) CN112329338B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112966399A (en) * 2021-04-15 2021-06-15 苏州大学张家港工业技术研究院 Pulse tube refrigerator working condition prediction method and system based on machine learning
CN113742999A (en) * 2021-08-12 2021-12-03 中国船舶重工集团公司第七一九研究所 Design method and device for printed circuit board type heat exchanger
CN114216246A (en) * 2021-12-16 2022-03-22 东软云科技有限公司 Air conditioning unit control method and device, storage medium and electronic equipment
CN116893614A (en) * 2023-06-06 2023-10-17 苏州优世达智能科技有限公司 Control method of amphibious unmanned ship based on multi-sensor fusion
CN117647932A (en) * 2024-01-25 2024-03-05 上海碳索能源服务股份有限公司 Method, system, terminal and medium for constructing cooling pump flow prediction model

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104089362A (en) * 2014-06-03 2014-10-08 杭州哲达科技股份有限公司 Cooling efficiency maximization method for cooling water system in central air-conditioner and control device
WO2015172560A1 (en) * 2014-05-16 2015-11-19 华南理工大学 Central air conditioner cooling load prediction method based on bp neural network
CN105160398A (en) * 2015-08-04 2015-12-16 莆田学院 Microbial fermentation optimizing method based on artificial fish school algorithm
CN107330510A (en) * 2017-06-30 2017-11-07 南京信息工程大学 Humidity sensor temperature compensation method based on AFSA BP neural networks
CN108089440A (en) * 2017-12-06 2018-05-29 北京百度网讯科技有限公司 Energy-saving control method and device
CN109084415A (en) * 2018-07-26 2018-12-25 杭州哲达节能科技有限公司 Central air-conditioning operating parameter optimization method based on artificial neural network and genetic algorithms
CN110222878A (en) * 2019-05-17 2019-09-10 广东工业大学 A kind of short-term load forecasting method based on artificial fish-swarm neural network
CN110285532A (en) * 2019-07-04 2019-09-27 中国工商银行股份有限公司 Method for controlling machine room air conditioner, apparatus and system based on artificial intelligence
CN110805997A (en) * 2019-11-14 2020-02-18 中金新源（天津）科技有限公司 Energy-saving control method for central air-conditioning system
CN110848896A (en) * 2019-11-12 2020-02-28 深圳孚沃德斯科技有限公司 Intelligent energy-saving control system and method for air conditioner cooling system based on neural network
CN111144543A (en) * 2019-12-30 2020-05-12 中国移动通信集团内蒙古有限公司 Data center air conditioner tail end temperature control method, device and medium

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2015172560A1 (en) * 2014-05-16 2015-11-19 华南理工大学 Central air conditioner cooling load prediction method based on bp neural network
CN104089362A (en) * 2014-06-03 2014-10-08 杭州哲达科技股份有限公司 Cooling efficiency maximization method for cooling water system in central air-conditioner and control device
CN105160398A (en) * 2015-08-04 2015-12-16 莆田学院 Microbial fermentation optimizing method based on artificial fish school algorithm
CN107330510A (en) * 2017-06-30 2017-11-07 南京信息工程大学 Humidity sensor temperature compensation method based on AFSA BP neural networks
CN108089440A (en) * 2017-12-06 2018-05-29 北京百度网讯科技有限公司 Energy-saving control method and device
CN109084415A (en) * 2018-07-26 2018-12-25 杭州哲达节能科技有限公司 Central air-conditioning operating parameter optimization method based on artificial neural network and genetic algorithms
CN110222878A (en) * 2019-05-17 2019-09-10 广东工业大学 A kind of short-term load forecasting method based on artificial fish-swarm neural network
CN110285532A (en) * 2019-07-04 2019-09-27 中国工商银行股份有限公司 Method for controlling machine room air conditioner, apparatus and system based on artificial intelligence
CN110848896A (en) * 2019-11-12 2020-02-28 深圳孚沃德斯科技有限公司 Intelligent energy-saving control system and method for air conditioner cooling system based on neural network
CN110805997A (en) * 2019-11-14 2020-02-18 中金新源（天津）科技有限公司 Energy-saving control method for central air-conditioning system
CN111144543A (en) * 2019-12-30 2020-05-12 中国移动通信集团内蒙古有限公司 Data center air conditioner tail end temperature control method, device and medium

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
LEAHY000: "Note——Neural Network and Deep Learning （1）[神经网络与深度学习学习笔记（1）]" *
牧野: "BP神经网络模型及梯度下降法" *
魏立新;张峻林;刘青松;: "基于改进人工鱼群算法的神经网络优化", 控制工程 *
龚波等: "一种改进人工鱼群算法对BP神经网络的优化研究", 《湖南科技大学学报(自然科学版)》 *

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112966399A (en) * 2021-04-15 2021-06-15 苏州大学张家港工业技术研究院 Pulse tube refrigerator working condition prediction method and system based on machine learning
CN112966399B (en) * 2021-04-15 2023-08-22 苏州大学张家港工业技术研究院 Pulse tube refrigerator working condition prediction method and system based on machine learning
CN113742999A (en) * 2021-08-12 2021-12-03 中国船舶重工集团公司第七一九研究所 Design method and device for printed circuit board type heat exchanger
CN113742999B (en) * 2021-08-12 2023-09-26 中国船舶重工集团公司第七一九研究所 Design method and device for printed circuit board type heat exchanger
CN114216246A (en) * 2021-12-16 2022-03-22 东软云科技有限公司 Air conditioning unit control method and device, storage medium and electronic equipment
CN114216246B (en) * 2021-12-16 2023-08-29 东软云科技有限公司 Air conditioning unit control method and device, storage medium and electronic equipment
CN116893614A (en) * 2023-06-06 2023-10-17 苏州优世达智能科技有限公司 Control method of amphibious unmanned ship based on multi-sensor fusion
CN116893614B (en) * 2023-06-06 2023-12-15 苏州优世达智能科技有限公司 Control method of amphibious unmanned ship based on multi-sensor fusion
CN117647932A (en) * 2024-01-25 2024-03-05 上海碳索能源服务股份有限公司 Method, system, terminal and medium for constructing cooling pump flow prediction model

Also Published As

Publication number Publication date
CN112329338B (en) 2023-07-28

Similar Documents

Publication Publication Date Title
CN112329338A (en) Cold source system control method and device based on fish swarm optimization BP neural network
CN111256294B (en) Model prediction-based optimization control method for combined operation of water chilling unit
CN108413567B (en) Central air conditioner cost-saving optimization method and system based on Internet of things
CN104134100B (en) A kind of energy-saving management system based on cloud computing
CN106920006A (en) A kind of subway station air conditioning energy consumption Forecasting Methodology based on ISOA LSSVM
CN113587414B (en) Air conditioner water system control system
WO2023030522A1 (en) Data center air conditioning system diagnosis method and apparatus
CN114383299B (en) Central air-conditioning system operation strategy optimization method based on big data and dynamic simulation
CN105303247A (en) Garden type hot and cold energy mixed application energy network regulation method and system
CN111144543A (en) Data center air conditioner tail end temperature control method, device and medium
CN112628956B (en) Water chilling unit load prediction control method and system based on edge cloud cooperative framework
CN113739365A (en) Central air-conditioning cold station group control energy-saving control method, device, equipment and storage medium
CN116266253A (en) Optimization control method, system and computer readable storage medium for air conditioner parameters
CN116398994B (en) Water chilling unit group control optimization method based on load prediction
CN113551373A (en) Data center air conditioner energy-saving control method based on federal reinforcement learning
CN112113314A (en) Real-time temperature data acquisition system and temperature adjusting method based on learning model
CN112413762A (en) Parameter optimization method and system for cooling water system of refrigerating room
CN115049141A (en) Control method and control device for refrigeration system and refrigeration system
CN114326987A (en) Refrigeration system control and model training method, device, equipment and storage medium
CN111787764B (en) Energy consumption optimization method and device for multi-split refrigerating unit, electronic equipment and storage medium
CN115264768A (en) Building air conditioner energy-saving control method and system based on Internet of things
CN116562111A (en) Data center energy saving method, device, system and storage medium
CN111043720B (en) Low-cost robustness adjustment strategy making method of refrigeration system under load uncertainty
CN114636212A (en) GRNN-based multi-chiller system operation control method
CN112944599A (en) Multi-parameter coupling control method and device of air conditioning system

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