CN113935556B - Temperature sensor optimal arrangement method based on DNA genetic algorithm - Google Patents
Temperature sensor optimal arrangement method based on DNA genetic algorithm Download PDFInfo
- Publication number
- CN113935556B CN113935556B CN202111536202.4A CN202111536202A CN113935556B CN 113935556 B CN113935556 B CN 113935556B CN 202111536202 A CN202111536202 A CN 202111536202A CN 113935556 B CN113935556 B CN 113935556B
- Authority
- CN
- China
- Prior art keywords
- optimization
- target
- temperature
- temperature sensor
- optimal
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
Abstract
The invention is suitable for the technical field of artificial intelligence and intelligent optimization calculation, and provides a temperature sensor optimal arrangement method based on a DNA genetic algorithm, which comprises the following steps: selecting point positions possibly arranged by the temperature sensor, and establishing a multi-objective optimization function by taking energy consumption measurement precision and energy consumption change sensitivity as optimization targets; establishing a multi-target expected value model, carrying out random simulation solving, and combining the multi-target optimization function to obtain a multi-target optimization model; randomly simulating by using a DNA genetic algorithm, and calculating the multi-objective optimization model; and obtaining target Pareto optimal solutions of the two optimization targets, repeatedly calculating to obtain a group of optimal solution sets, using the optimal solution sets as an input set of a deep learning algorithm, and searching for a better solution through the deep learning algorithm.
Description
Technical Field
The invention belongs to the technical field of artificial intelligence and intelligent optimization calculation, and particularly relates to a temperature sensor optimal arrangement method based on a DNA genetic algorithm.
Background
Under the background of smart cities, energy consumption supervision and energy-saving optimization of large buildings are leading to intelligent roads. By means of information technologies such as the Internet of things, cloud computing and big data, a more comprehensive and accurate data base can be provided, and the method has extremely important guiding significance for development of building energy-saving work in China. The key to obtaining a comprehensive and accurate data base is sensor acquisition, which involves a core problem, namely how to use the minimum number of sensors to optimize the arrangement of the sensors since the sensors acquire enough information.
The point location capable of arranging the sensors on the actual building structure of the sensor optimal arrangement problem is a research object, belongs to a discrete model, and is an integer programming combined optimization problem. If the number of the sensors is s and the number of the candidate positions is m, the total number isAn arrangement scheme, which is calculated if the enumeration method is adopted to calculate the optimized objective functionSecondly, for a structure with a small scale, the calculation method is acceptable, and the calculation method is difficult to solve for a large building structure. Especially, the temperature sensor is typical, the arrangement positions of the temperature sensor are many, and the temperature sensor is an important index for energy saving and consumption reduction, for example, the indoor comfort can be kept without cold summer and hot winter through intelligent and accurate temperature control, which is the key of energy saving. The solving of the problems is to establish a random integer programming model, in order to avoid dimension explosion, intelligent algorithms such as a genetic algorithm and the like are needed to be adopted for solving, and a DNA genetic algorithm is a genetic algorithm which is good in integer programming solving performance.
Temperature spatial distribution is difficult to measure directly, and usually, a plurality of single-point or ultrasonic temperature sensors are researched to form a sensor temperature measurement array, so that temperature field information is obtained by adopting an interpolation or fitting mode. The data redundancy of the sensors is caused by the excessive arrangement of the sensors, so the cost is increased; the insufficient number of the arranged sensors cannot sufficiently describe the spatial distribution information of the temperature field, and has a great influence on the accuracy of temperature control. Therefore, to obtain the temperature field distribution, the number of sensors and the measuring point positions are optimized, and the temperature field reconstruction is carried out by using the number and position information. In addition, the temperature change sensitivity of the sensor is also a key measure for the rational arrangement of the sensor.
The prior art is provided with a DNA genetic algorithm and a deep learning method aiming at an optimal arrangement method of a sensor, and the deep learning method is a tool for solving big data reasoning along with the development of artificial intelligence. However, for the optimal arrangement of the temperature sensors, the optimal objective function for performing the optimal arrangement of the sensors cannot be obtained by the current DNA genetic algorithm and deep learning, so that the improvement of the temperature measurement accuracy and the temperature change sensitivity of the temperature sensors is not accurate enough, and the energy saving and consumption reduction are not facilitated.
Disclosure of Invention
In view of the above problems, the present invention aims to provide a deep learning temperature sensor optimal arrangement method based on a DNA genetic algorithm, and aims to solve the technical problem that the existing optimal arrangement model of the temperature sensor is not accurate enough.
The invention adopts the following technical scheme:
the deep learning temperature sensor optimal arrangement method based on the DNA genetic algorithm comprises the following steps:
s1, selecting possible point positions of the temperature sensors, and establishing a multi-objective optimization function by taking the energy consumption measurement precision and the energy consumption change sensitivity as optimization targets;
s2, establishing a multi-target expected value model, carrying out random simulation solving, and combining the multi-target optimization function to obtain a multi-target optimization model;
s3, calculating the multi-objective optimization model by random simulation of a DNA genetic algorithm;
and step S4, obtaining target Pareto optimal solutions of two optimization targets, repeatedly calculating to obtain a group of optimal solution sets, using the optimal solution sets as an input set of a deep learning algorithm, and searching for a better solution through the deep learning algorithm.
The invention has the beneficial effects that: the invention provides a more reasonable temperature sensor arrangement measure function, a multi-objective optimization model taking energy consumption measurement precision and energy consumption change sensitivity as two optimization targets is formed, then a DNA genetic algorithm suitable for solving the integer programming problem is adopted for solving, the solved result provides a sample for deep learning to obtain a better solution, and the deep learning precision and efficiency can be further improved; the scheme of the invention can improve the temperature measurement precision and the temperature change sensitivity of the temperature sensor, and provides a new alternative for intelligently exploring the optimal arrangement of the temperature sensor and saving energy and reducing consumption.
Drawings
FIG. 1 is a flow chart of a deep learning temperature sensor optimal arrangement method based on a DNA genetic algorithm provided by an embodiment of the invention;
FIG. 2 is a flowchart illustrating an implementation of step S1 according to an embodiment of the present invention;
fig. 3 is a flowchart of an implementation of step S2 according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
In order to explain the technical means of the present invention, the following description will be given by way of specific examples.
Fig. 1 shows a flow of a deep learning temperature sensor optimal arrangement method based on a DNA genetic algorithm according to an embodiment of the present invention, and only the parts related to the embodiment of the present invention are shown for convenience of description.
As shown in fig. 1, the deep learning temperature sensor optimal arrangement method based on the DNA genetic algorithm provided in this embodiment includes the following steps:
and S1, selecting possible point positions of the temperature sensors, and establishing a multi-objective optimization function by taking the energy consumption measurement precision and the energy consumption change sensitivity as optimization targets.
The step realizes the establishment of a multi-objective optimization function, the multi-objective optimization function considers the energy consumption measurement precision and the energy consumption change sensitivity, and the specific realization process is as follows:
and S11, selecting a point where the temperature sensor can be arranged.
Generally, a floor plan is established according to an actual arrangement situation, positions limited by physical conditions need to be removed, and the positions are uniformly distributed on the floor plan according to a minimum distance to generate N suitable positions, so that the final problem is that N temperature sensors are arranged at the N positions, and the purposes of good effect, energy conservation and money saving are achieved.
And S12, reconstructing the temperature field.
RBF (radial basis function) is generally used to solve the problem of multivariate interpolation and has been shown to approximate arbitrary functions with arbitrary accuracy, suitable for reconstruction of temperature fields.
N temperature sensors are arranged in a two-dimensional space, wherein the temperature of the jth measuring pointIn (1),andrespectively possible sensor placements are plotted on the abscissa and ordinate. The jth sensor may be centered on a radial basis interpolation, where the radial basis function may be defined as:wherein, in the step (A),is a shape parameter of the radial basis function, is related to the arrangement position of the temperature sensor,in the formulaThe maximum distance between the points is measured by the sensor.
In order to obtain the distribution characteristics of the temperature in the two-dimensional space, the measuring points are interpolated by adopting a radial basis function, namelyIn the formula (I), wherein,is a temperature field reconstructed using RBF;are parameters to be determined and can be derived from this formula:wherein, in the step (A),o represents a 0 matrix.
S13, constructing a precision objective functionTo describe the accuracy of the reconstructed temperature field; n is the total number of temperature sensors, j is the jth temperature sensor,the measured point temperature of the jth temperature sensor,is the temperature field of the jth temperature sensor, f1Is a precision objective function.
In order to improve the accuracy of temperature field reconstruction as much as possible, an accuracy objective function is required to describe the accuracy of the reconstructed temperature field, and the smaller the objective value, the better.
S14, constructing a sensitivity objective functionFor the description ofReconstructing the sensitivity of the temperature field to the temperature difference change, n being the total number of temperature sensors, j being the j temperature sensor, f2In order to be the objective function of the sensitivity,is composed ofThe smaller the target value is, the better, and。
in order to improve the sensitivity of the temperature field reconstruction to temperature difference changes as much as possible, an objective function is required to be constructed to describe the sensitivity of the reconstructed temperature field to temperature difference changes, and the smaller the objective value, the better.
Finally f is obtained according to the above1、f2The multi-objective optimization function is to take f1、f2Is measured.
And S2, establishing a multi-target expected value model, performing random simulation solution, and combining the multi-target optimization function to obtain a multi-target optimization model.
The method specifically comprises the following steps:
and S21, establishing a multi-target expected value model.
In the random-integer programming expectation-value model, if each decision variable is required to be an integer, it can be regarded as a random-integer programming expectation-value model, let the decision vector be x,is a random vector of the number of bits,is an objective function, andis a set of random constraint functions, p is the number of constraint functions,e denotes the expectation operator, if and only ifTo solveIs feasible. If for each feasible solutionSatisfy the requirement ofThis means a feasible solutionIs the optimal solution for this expectation model.
In many cases, when many objective functions are to be considered, in this embodiment two objective functions. If the decision maker wishes to maximize the expectation of these objective functions. The present embodiment can establish the following multi-target expectation value model.
And S22, solving the multi-target expected value model through random simulation to obtain a final multi-target optimization model.
Is provided withIs in a probability spaceIs/are as followsThe dimensions of the random vector are then calculated,is a sample space, A is an event field, isThe number of generations is determined by the number of generations,is a probability measure.Is a measurable function, thenIs a random variable. The detailed steps are as follows: is provided with(ii) a According to probability measureFromIn the production of samples;③(ii) a Fourthly, repeating the step III and the step III for N times; fifthly。
And S3, calculating the multi-objective optimization model by random simulation of a DNA genetic algorithm.
The specific process using the DNA genetic algorithm is as follows:
s31, inputting parameters, wherein the parameters comprise population scale, cross probability, mutation probability and inversion probability;
inputting parametersWhereinIn order to be of the population scale,in order to be a cross-over probability,the probability of the variation is the probability of the variation,is the probability of inversion;
s32, initial GenerationThe DNA strands forming the initial populationAnd verifying the feasibility of the DNA chains by using a random simulation technology;
s33, calculating the fitness: translating the codon of each DNA chain in the population into a parameter value according to a cipher table, and then calculating the fitness of the individual by using a random simulation technology;
s34, selecting: from DNA strand populations with a certain probabilitySelecting m DNA chain individuals as parents for breeding offspring, verifying the feasibility of the offspring by using a random simulation technology, and adding new individuals meeting the feasibility into the next generation;
S35, carrying out crossing, mutation and inversion operation on the DNA chain, wherein the feasibility of the offspring is still tested by using random simulation;
The multi-objective optimization problem does not generally have a so-called absolute optimal solution, but 'compromises' among all objective functions and gives a solution of the problem in the form of a Pareto optimal solution set, and a final solution is selected from the Pareto optimal solution set by a decision maker according to actual conditions. The objective function vector corresponding to the Pareto optimal solution set forms a space curved surface, and the optimal solution can be found according to the Pareto front surface.
And step S4, obtaining target Pareto optimal solutions of two optimization targets, repeatedly calculating to obtain a group of optimal solution sets, using the optimal solution sets as an input set of a deep learning algorithm, and searching for a better solution through the deep learning algorithm.
S41, calculating to obtain target Pareto optimal solutions of the two optimization targets through the step S3, and repeatedly calculating the target Pareto optimal solutions of the two optimization targets to form a group of optimal solution sets;
s42, taking the optimal solution set as a training set, a verification set and a test set of a deep learning algorithm;
s43, deep learning training and verification are carried out;
and S44, reasoning the test set by using the trained deep learning model to obtain a more optimal solution arrangement mode.
The deep learning method is the prior art, and the detailed description of the specific calculation process is omitted here.
In conclusion, the energy consumption measurement precision and the energy consumption change sensitivity are used as two optimization targets to establish the multi-target optimization model, then the DNA genetic algorithm is adopted to solve, the solved result provides a sample for deep learning to obtain a better solution, the deep learning precision and efficiency can be further improved, and the temperature measurement precision and the temperature change sensitivity of the temperature sensor can be improved.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (3)
1. A temperature sensor optimal arrangement method based on a DNA genetic algorithm is characterized by comprising the following steps:
s1, selecting possible point positions of the temperature sensors, and establishing a multi-objective optimization function by taking the energy consumption measurement precision and the energy consumption change sensitivity as optimization targets;
s2, establishing a multi-target expected value model, carrying out random simulation solving, and combining the multi-target optimization function to obtain a multi-target optimization model;
s3, calculating the multi-objective optimization model by random simulation of a DNA genetic algorithm;
s4, obtaining target Pareto optimal solutions of two optimization targets, repeatedly calculating to obtain a group of optimal solution sets serving as an input set of a deep learning algorithm, and searching for a better solution through the deep learning algorithm;
the step S1 specifically includes:
s11, selecting possible arrangement points of the temperature sensors;
s12, reconstructing a temperature field;
s13, constructing a precision objective functionTo describe the accuracy of the reconstructed temperature field; n is the total number of temperature sensors, j is the jth temperature sensor,the measured point temperature of the jth temperature sensor,is the temperature field of the jth temperature sensor, f1Is a precision objective function;
s14, constructing a sensitivity objective functionDescribing the sensitivity of the reconstructed temperature field to the temperature difference change, n is the total number of temperature sensors, j is the jth temperature sensor, f2In order to be the objective function of the sensitivity,is composed ofThe mean value of (a);
the step S2 specifically includes:
s21, establishing a multi-target expected value modelWherein the content of the first and second substances,is a set of random constraint functions, p is the number of constraint functions, x is the decision vector,is a random vector, E represents the expectation operator;
s22, solving the multi-target expected value model through stochastic simulation to obtain a final multi-target optimization model
2. The method for optimal placement of temperature sensors based on DNA genetic algorithm as claimed in claim 1, wherein said step S3 specifically comprises:
s31, inputting parameters, wherein the parameters comprise population scale, cross probability, mutation probability and inversion probability;
s32, initially generating a DNA chain to form an initial population, and verifying the feasibility of the DNA chain by random simulation;
s33, translating the codon of each DNA chain in the population into a parameter value according to a cipher table, and then calculating the fitness of the individual by using random simulation;
s34, selecting DNA chain individuals from the DNA chain population according to a certain probability, using the DNA chain individuals as parents for breeding offspring, verifying the feasibility of the offspring by using random simulation, and adding new individuals meeting the feasibility into the next generation;
s35, carrying out crossing, mutation and inversion operation on the DNA chain, wherein the feasibility of the offspring is still tested by using random simulation;
s36, repeating S33-S35 until meeting the convergence condition.
3. The method for optimal placement of temperature sensors based on DNA genetic algorithm as claimed in claim 2, wherein said step S4 specifically comprises:
s41, calculating to obtain target Pareto optimal solutions of the two optimization targets through the step S3, and repeatedly calculating the target Pareto optimal solutions of the two optimization targets to form a group of optimal solution sets;
s42, taking the optimal solution set as a training set, a verification set and a test set of a deep learning algorithm;
s43, deep learning training and verification are carried out;
and S44, reasoning the test set by using the trained deep learning model to obtain a more optimal solution arrangement mode.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111536202.4A CN113935556B (en) | 2021-12-16 | 2021-12-16 | Temperature sensor optimal arrangement method based on DNA genetic algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111536202.4A CN113935556B (en) | 2021-12-16 | 2021-12-16 | Temperature sensor optimal arrangement method based on DNA genetic algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113935556A CN113935556A (en) | 2022-01-14 |
CN113935556B true CN113935556B (en) | 2022-03-22 |
Family
ID=79289071
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111536202.4A Active CN113935556B (en) | 2021-12-16 | 2021-12-16 | Temperature sensor optimal arrangement method based on DNA genetic algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113935556B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116992780B (en) * | 2023-09-25 | 2024-01-02 | 中南大学 | Temperature sensor arrangement method for digital electrolytic tank |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6006604A (en) * | 1997-12-23 | 1999-12-28 | Simmonds Precision Products, Inc. | Probe placement using genetic algorithm analysis |
CN107391789A (en) * | 2017-06-12 | 2017-11-24 | 西安理工大学 | Silicon melt reconstructing temperature field method based on free surface measured temperature and characteristic function interpolation |
CN111368970A (en) * | 2020-02-17 | 2020-07-03 | 哈尔滨工业大学 | Sensor optimal arrangement method based on deep reinforcement learning |
CN112989529A (en) * | 2021-02-05 | 2021-06-18 | 河北农业大学 | Henhouse temperature sensor arrangement optimization method based on genetic algorithm |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106875071A (en) * | 2017-03-31 | 2017-06-20 | 华南理工大学 | A kind of workshop appliance multiple-objection optimization layout method of self-adapted genetic algorithm |
CN112181867B (en) * | 2020-09-29 | 2022-07-26 | 西安电子科技大学 | On-chip network memory controller layout method based on multi-target genetic algorithm |
-
2021
- 2021-12-16 CN CN202111536202.4A patent/CN113935556B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6006604A (en) * | 1997-12-23 | 1999-12-28 | Simmonds Precision Products, Inc. | Probe placement using genetic algorithm analysis |
CN107391789A (en) * | 2017-06-12 | 2017-11-24 | 西安理工大学 | Silicon melt reconstructing temperature field method based on free surface measured temperature and characteristic function interpolation |
CN111368970A (en) * | 2020-02-17 | 2020-07-03 | 哈尔滨工业大学 | Sensor optimal arrangement method based on deep reinforcement learning |
CN112989529A (en) * | 2021-02-05 | 2021-06-18 | 河北农业大学 | Henhouse temperature sensor arrangement optimization method based on genetic algorithm |
Non-Patent Citations (3)
Title |
---|
基于多目标遗传算法的传感器优化布点研究;张连振等;《工程力学》;20070430;第24卷(第4期);第168-173页 * |
基于改进DNA遗传算法的微电网DG选址定容研究;宋雷震等;《黑龙江工业学院学报》;20201031;第20卷(第10期);第81-87页 * |
水稻浸种催芽箱温度传感器优化配置——基于遗传算的发;李纯思等;《农机化研究》;20170630(第6期);第27-32页 * |
Also Published As
Publication number | Publication date |
---|---|
CN113935556A (en) | 2022-01-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103077288B (en) | Towards hard measurement and the formula decision-making technique thereof of the multicomponent alloy material of small sample test figure | |
CN109684714B (en) | Building design method based on machine learning and BIM technology | |
CN104636985A (en) | Method for predicting radio disturbance of electric transmission line by using improved BP (back propagation) neural network | |
CN108564205A (en) | A kind of load model and parameter identification optimization method based on measured data | |
CN110232461A (en) | More interconnection vector machine water quality prediction methods based on quantum genetic algorithm optimization | |
CN110083125A (en) | A kind of machine tool thermal error modeling method based on deep learning | |
CN108805346A (en) | A kind of hot continuous rolling force forecasting method based on more hidden layer extreme learning machines | |
Yan | Research on green suppliers' evaluation based on AHP & genetic algorithm | |
CN113935556B (en) | Temperature sensor optimal arrangement method based on DNA genetic algorithm | |
CN103530700B (en) | Urban distribution network saturation loading Comprehensive Prediction Method | |
CN103473465B (en) | Land resource spatial configuration optimal method based on multiple target artificial immune system | |
Liang et al. | An enhanced slime mould algorithm and its application for digital IIR filter design | |
CN113591368A (en) | Comprehensive energy system multi-energy load prediction method and system | |
CN115310348A (en) | Stacking-based grouting amount integrated agent prediction model and prediction method | |
CN106295877B (en) | Method for predicting electric energy consumption of smart power grid | |
CN109408896B (en) | Multi-element intelligent real-time monitoring method for anaerobic sewage treatment gas production | |
Fan | Research on deep learning energy consumption prediction based on generating confrontation network | |
Ping et al. | A mixed isomorphism approach for kinematic structure enumeration graphs based on intelligent design and manufacturing | |
CN110033118A (en) | Elastomeric network modeling and the blower multiobjective optimization control method based on genetic algorithm | |
CN111126827A (en) | Input-output accounting model construction method based on BP artificial neural network | |
CN113762591B (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM countermeasure learning | |
CN103778327A (en) | Curve fitting method | |
Du et al. | A modified niche genetic algorithm based on evolution gradient and its simulation analysis | |
Gu | Innovative Sensor Design Method of Industrial Ceramic Product Modeling Based on Immune Optimization Algorithm | |
Lu et al. | Application of GA optimized wavelet neural networks for carrying capacity of water resources prediction |
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 |