CN114167726A - Thermal power plant dust real-time monitoring system and method based on energy conservation optimization - Google Patents
Thermal power plant dust real-time monitoring system and method based on energy conservation optimization Download PDFInfo
- Publication number
- CN114167726A CN114167726A CN202111463373.9A CN202111463373A CN114167726A CN 114167726 A CN114167726 A CN 114167726A CN 202111463373 A CN202111463373 A CN 202111463373A CN 114167726 A CN114167726 A CN 114167726A
- Authority
- CN
- China
- Prior art keywords
- dust concentration
- dust
- power plant
- thermal power
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000000428 dust Substances 0.000 title claims abstract description 154
- 238000000034 method Methods 0.000 title claims abstract description 59
- 238000012544 monitoring process Methods 0.000 title claims abstract description 50
- 238000005457 optimization Methods 0.000 title claims abstract description 38
- 238000004134 energy conservation Methods 0.000 title description 4
- 230000002068 genetic effect Effects 0.000 claims abstract description 28
- 230000007613 environmental effect Effects 0.000 claims abstract description 26
- 238000013528 artificial neural network Methods 0.000 claims abstract description 22
- 230000005540 biological transmission Effects 0.000 claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000010276 construction Methods 0.000 claims abstract description 9
- 238000001514 detection method Methods 0.000 claims description 17
- 230000035772 mutation Effects 0.000 claims description 15
- 230000004927 fusion Effects 0.000 claims description 13
- 238000007664 blowing Methods 0.000 claims description 6
- 238000003062 neural network model Methods 0.000 claims description 5
- 230000003287 optical effect Effects 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000004140 cleaning Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 44
- 239000002245 particle Substances 0.000 description 20
- 230000003044 adaptive effect Effects 0.000 description 11
- 238000005265 energy consumption Methods 0.000 description 9
- 108090000623 proteins and genes Proteins 0.000 description 8
- 230000002829 reductive effect Effects 0.000 description 8
- 241000854291 Dianthus carthusianorum Species 0.000 description 6
- 238000004891 communication Methods 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 6
- 230000006855 networking Effects 0.000 description 6
- 230000008033 biological extinction Effects 0.000 description 5
- 238000004880 explosion Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000009826 distribution Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 239000002817 coal dust Substances 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 230000005059 dormancy Effects 0.000 description 3
- 238000010521 absorption reaction Methods 0.000 description 2
- 238000000149 argon plasma sintering Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 210000000349 chromosome Anatomy 0.000 description 2
- 239000003245 coal Substances 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 238000000097 high energy electron diffraction Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 206010035653 pneumoconiosis Diseases 0.000 description 2
- 238000010248 power generation Methods 0.000 description 2
- 230000007958 sleep Effects 0.000 description 2
- 230000004083 survival effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 108700028369 Alleles Proteins 0.000 description 1
- 206010016936 Folliculitis Diseases 0.000 description 1
- 208000028571 Occupational disease Diseases 0.000 description 1
- 206010036590 Premature baby Diseases 0.000 description 1
- 206010049590 Upper respiratory tract inflammation Diseases 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000005315 distribution function Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000010353 genetic engineering Methods 0.000 description 1
- 230000008303 genetic mechanism Effects 0.000 description 1
- 239000008187 granular material Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000012528 membrane Substances 0.000 description 1
- 230000004630 mental health Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001590 oxidative effect Effects 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000002028 premature Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000005180 public health Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 239000012798 spherical particle Substances 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The application discloses a thermal power plant dust real-time monitoring system and method based on energy-saving optimization, wherein the system comprises a data acquisition module, a model construction module and a monitoring module; the data acquisition module is used for acquiring dust concentration information and environmental information of the thermal power plant; the model construction module is used for establishing a dust concentration prediction model based on a BP neural network and optimizing the dust concentration prediction model based on an improved genetic algorithm; the monitoring module is used for carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model; the data acquisition module, the model building module and the monitoring module are in data transmission in a wireless connection mode. The method and the device perform multi-level and multi-level optimization processing on data from a plurality of sensors, establish a dust concentration prediction model by improving a genetic algorithm and a BP neural network, comprehensively consider dust concentration and dust environmental factors, and perform prediction and early warning.
Description
Technical Field
The invention relates to the field of monitoring and judging productive dust in a working place of a thermal power plant, in particular to a thermal power plant dust real-time monitoring system and method based on energy-saving optimization.
Background
In China, the energy consumption is mainly coal, and wind power generation and solar power generation technologies are not mature, so that a thermal power plant also plays an important role. After raw coal enters a power plant, a large amount of dust is generated in the whole process through handling, transporting, screening, crushing and other treatment processes. If the coal dust is allowed to spread all around, serious harm is caused.
On one hand, the diffusion of coal dust can accelerate the abrasion and aging of mechanical equipment or the failure of parts, endanger the health of workers, cause diseases such as upper respiratory tract inflammation, folliculitis and pneumoconiosis, and seriously damage the physical and mental health of workers. Data show that more than 75% of occupational diseases in China are pneumoconiosis, and dust hazard becomes a more serious problem in the current global public health problem. On the other hand, a considerable amount of coal dust causes a rather serious problem of dust explosion. The general definition of dust explosion is: the process of releasing energy violently occurs by the rapid combustion of fine solid particles, liquid droplets, suspended in a gas, here an oxidizing gas, generally air, in a confined space. The explosion also produces a secondary explosion in most cases, which is more destructive. According to statistics, dust explosion accidents happen at least together every day on average all over the world. Further verifies the urgency of the dust explosion-proof industry construction.
Therefore, how to better monitor the change of the dust concentration in the production environment so as to perform timely and accurate dust control measures becomes a problem which needs to be solved urgently. In the traditional dust control, the dust concentration condition is detected in a manual sampling inspection mode, and a series of problems of data failure, narrow coverage, long detection period, incapability of continuous monitoring and the like exist. With the progress of a machine learning algorithm and the breakthrough development of the technology of the internet of things in recent years, the ability of people to sense the surrounding environment and judge the mode is remarkably improved, so that the establishment of an effective dust environment monitoring system is an effective solution.
Disclosure of Invention
The application provides a thermal power plant dust real-time monitoring system and method based on energy conservation optimization, which realize quick information acquisition of dust concentration through a wireless network sensor, realize networking of dust concentration information acquired at each point through a network technology, thereby forming a dust environment real-time monitoring system and realizing prediction and early warning of thermal power plant dust concentration change.
In order to achieve the above purpose, the present application provides the following solutions:
a thermal power plant dust real-time monitoring system based on energy conservation optimization comprises: the system comprises a data acquisition module, a model construction module and a monitoring module;
the data acquisition module is used for acquiring dust concentration information and environmental information of the thermal power plant;
the model construction module is used for establishing a dust concentration prediction model based on a BP neural network and optimizing the dust concentration prediction model based on an improved genetic algorithm;
the monitoring module is used for carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model;
the data acquisition module, the model building module and the monitoring module are in data transmission in a wireless connection mode.
Preferably, the data acquisition module comprises a detection unit and a sensor unit;
the detection unit is used for detecting the dust concentration by a light transmission method;
the sensor unit is used for receiving the dust concentration, collecting the environmental information and sending the dust concentration and the environmental information to the monitoring module;
the detection unit and the sensor unit are in wireless connection for data transmission.
Preferably, the sensor unit performs data fusion and optimization processing on the dust concentration information and the environmental information by using a support function.
Preferably, the detection unit includes: the device comprises a laser, a photoelectric detector, a signal processing circuit, a singlechip, a computer, a power supply and an air blowing device;
the laser is used for emitting laser beams;
the photoelectric detector is used for receiving the laser beam in the detected dust area and converting the optical signal of the laser beam into an electric signal;
the air blowing device is used for cleaning a window of the laser and a window of the photoelectric detector;
the signal processing circuit is used for receiving the electric signal and converting the electric signal into a digital signal;
the single chip microcomputer is used for preprocessing the digital signal to obtain the dust concentration data and sending the dust concentration data to the computer;
the computer displays the received dust concentration data by using LabVIEW software.
Preferably, when the dust concentration is detected, an auxiliary detector is arranged to receive ambient light.
Preferably, the optimization method of the dust concentration prediction model comprises the following steps: optimizing the initial weight of the BP neural network based on an improved genetic algorithm, and then performing back propagation operation adjustment to obtain a generalized neural network model serving as the optimized dust concentration prediction model; the improved genetic algorithm is an improved self-adaptive adjusting method based on cross probability and mutation probability.
Preferably, the dust concentration prediction model is further optimized by the dust concentration information and the environmental information.
The application also discloses a thermal power plant dust real-time monitoring method based on energy-saving optimization, which is used for collecting thermal power plant dust concentration information and environmental information;
establishing a dust concentration prediction model based on a BP neural network, and optimizing the dust concentration prediction model based on an improved genetic algorithm;
and carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model.
Preferably, the method for obtaining the dust concentration comprises the following steps: and continuously detecting the dust by using a light transmission method to obtain the dust concentration.
Preferably, the dust concentration information and the dust environment information are subjected to data fusion and optimization processing by constructing a support function.
The beneficial effect of this application does:
the application discloses a thermal power plant dust real-time monitoring system and method based on energy-saving optimization, the dust concentration is rapidly acquired through a wireless network sensor, multi-level and multi-level optimization processing is carried out on data from a plurality of sensors, a dust concentration prediction model is established through improving a genetic algorithm and a BP neural network, and dust concentration information acquired at each point is networked, so that a dust environment real-time monitoring system is formed, and the dust concentration change is predicted and early-warned by comprehensively considering factors such as dust characteristics, dust release source intensity, environment humidity, wind speed and wind direction. The dust environment monitoring system established by the application is more suitable for the special operation environment of the thermal power plant, realizes real-time dust concentration monitoring and early warning in a larger range under the condition of not investing a large amount of manpower, and provides an effective means for dust prevention and control and operation environment monitoring.
Drawings
In order to more clearly illustrate the technical solution of the present application, the drawings needed to be used in the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to these drawings without inventive exercise.
Fig. 1 is a schematic structural diagram of a thermal power plant dust real-time monitoring system based on energy saving optimization according to a first embodiment of the present application;
fig. 2 is a schematic structural diagram of a detection unit according to a first embodiment of the present application;
fig. 3 is a schematic diagram of a signal processing flow according to a first embodiment of the present application;
FIG. 4 is a diagram illustrating an encoding method according to a first embodiment of the present application;
fig. 5 is a schematic flow chart of a thermal power plant dust real-time monitoring method based on energy-saving optimization in the second embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, the present application is described in further detail with reference to the accompanying drawings and the detailed description.
Example 1: as shown in fig. 1, a thermal power plant dust real-time monitoring system based on energy-saving optimization comprises a data acquisition module, a model construction module and a monitoring module;
the data acquisition module is used for acquiring dust concentration information and environmental information of the thermal power plant;
the model construction module is used for establishing a dust concentration prediction model based on a BP neural network and optimizing the dust concentration prediction model based on an improved genetic algorithm;
the monitoring module is used for carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model;
and the data acquisition module, the model building module and the monitoring module are in data transmission in a wireless connection mode.
Specifically, the data acquisition module comprises a detection unit and a sensor unit;
the detection unit detects the dust concentration by using a light transmission method;
the specific process of the method for continuously detecting the dust concentration in the dust environment monitoring system comprises the following steps:
the invention selects a light transmission method to continuously detect high-concentration dust in the environment of the thermal power plant, and designs a dust environment monitoring system and a signal circuit based on the environmental characteristics of the thermal power plant.
At present, the detection methods for dust concentration mainly include a filter membrane weighing method, a ray method, an ultrasonic attenuation method, a charge induction method, a capacitance method, a light scattering method, a light transmission method and the like, but the method generally has the limitations that online monitoring cannot be carried out or the concentration of the dust is low, and the method is generally applied to detection of respirable dust PM10/2.5 within the range of 0.1-1000 mg/m 3. The light transmission method can accurately measure low-concentration dust by combining with scattered light, and is more suitable for detecting high-concentration dust simply according to a transmitted light signal, so that the requirement of monitoring explosive high-concentration dust is met.
The light transmission method, also called extinction method or turbidimetric method, is based on the well-known Lambert-Beer theorem and the basic principle is that when a beam of light passes through a medium containing particles, the intensity of the transmitted light passing through the medium is attenuated due to the scattering and absorption effects of the particles, and the attenuation degree is related to the size, concentration, optical path and extinction coefficient of the particles, and the relation is
I＝I0exp(-τl) (1)
In the formula: i-transmitted light intensity;
i0-incident light intensity;
τ -turbidity;
l-optical path length.
Assuming that the particles to be detected in a unit volume are N monodisperse spherical particle systems with a light-receiving area a, the turbidity caused by light scattering and absorption is
Wherein D is the particle diameter; qext is the extinction coefficient, which is related to the wavelength λ of the incident light, the particle diameter D, the refractive index m of the particle relative to the surrounding medium. The calculation formula of the particle number concentration obtained by simultaneous formulas (1) and (2) is
When the particle density ρ is known, the particle mass concentration formula can be obtained from equation (3)
In practical situation, the tested particles are almost impossible to be monodisperse particle systems with the same particle size, but are polydisperse particle systems distributed in a certain size range, and assuming that the size distribution function is N (D), the converted calculation formula is
Wherein, the incident light intensity I0 is obtained by measurement when no dust exists; the optical path l is the length of the laser passing through the dust, namely the distance between the laser and the detector; the particle size distribution is obtained by measuring the measured dust by a laser particle size analyzer in advance; the extinction coefficient Qext can be obtained by the following equation:
where α is a dimensionless particle size parameter and is a function of particle size D and incident light wavelength λ, α ═ π D/λ, an、bnFor the Mie scattering coefficient, it can be obtained by the following formula:
wherein psin(z)、ξn(z) is the Ricitti-Bessel function, whichThese are functions of the Bessel function of the semiinteger order and the Hankel function of the second kind (z may be α or m α), and can be obtained by the following formula:
wherein:
in the formula, Nn+1/2(z) is a Neumann function, which may also be referred to as a Bessel function of the second type. The extinction coefficient Qext can be obtained by equations (6) to (11). As can be seen from the equations (1) to (11), when the measured particles and the test system determine that ρ, D, l and Qext are constants, the mass concentration of the dust is a function ln (I) of the transmitted light intensity and the incident light intensity0I) is linear, so the dust concentration can be solved by detecting the transmitted light intensity.
As shown in fig. 2, the detection unit includes: the device comprises a laser, a photoelectric detector, a signal processing circuit, a singlechip, a computer, a power supply and an air blowing device;
according to the principle of detecting the dust concentration by a transmission method, a laser beam emitted by a system penetrates through the detected dust, the transmission light intensity is detected by a photoelectric detector, then a weak signal converted by the photoelectric detector is amplified, filtered and subjected to A/D conversion, a single chip microcomputer is used for preprocessing the converted digital signal and then transmitting the preprocessed digital signal to a PC (personal computer) through RS485 communication, and the received data is processed and displayed by the upper computer through LabVIEW software. In addition, because laser instrument window and photoelectric detector window are exposed in the dusty air current, can the different degree adhesion dust granule, cause the influence to the test result, consequently added the air-blowing device and be used for cleaning the window.
Specifically, when the dust concentration is detected, an auxiliary detector is arranged to receive ambient light.
As shown in fig. 3, considering that the ambient light has a certain influence on the photodiode in addition to the laser light source, in order to eliminate the influence of the ambient light, another photodiode is added in the system as an auxiliary detector for receiving the ambient light, and then the differential processing is performed with the main detector through a differential circuit.
Specifically, the sensor unit is used for receiving the dust concentration, collecting dust environment information and sending the dust concentration and the dust environment information to the monitoring module;
specifically, the detection unit and the sensor unit are in wireless connection for data transmission.
Specifically, the sensor unit determines node location and coverage control based on an improved genetic algorithm.
In order to solve the problems of coverage redundancy and excessively fast node energy consumption under the condition of random deployment, sensor nodes are divided into two modes of dormancy and work on the basis of ensuring the network coverage quality, the combination of the work nodes in the network is called a coverage set, and the optimal coverage control aims to select a group of optimal nodes from all the randomly deployed sensor nodes to work, so that the sensor network has the maximized coverage rate and the minimized number of the work nodes. The weighting method is a classical algorithm in a combinatorial optimization problem that can convert a multi-objective function into a single objective function. In practical application, the optimal coverage of the wireless sensor network is a multi-target task, the multi-target task can be converted into a single-target task function by a weighting method, and meanwhile, the weight of each target function can be considered according to practical conditions to carry out personalized setting.
(1) Solution target and fitness selection
When Rc is 2Rs, the wireless sensor network may be considered to be fully connected under the premise of sufficient coverage, and connectivity may not be considered in the objective function. Therefore, the objective function of the optimal coverage control of the wireless sensor network is the coverage set with the maximized coverage rate and the minimum number of working nodes. Here, subsets are defined
The solution objective is two objective functions, namely the sub objective functions:
target 1: coverage ratio Pcov(S') max, i.e.:
target 2: the number of working nodes in the subset S' is the minimum, that is, the node dormancy rate is the maximum, that is:
maxf2＝1-|S'|/|S| (13)
the fitness function is equivalent to the survival ability of species in biology, the fitness function in this chapter is obtained by solving an objective formula (12) and an objective formula (13), and according to a weighting method, the total objective function is defined as the weighted sum of two objective functions, which is expressed as:
maxf＝ω1f1+ω2f2 (14)
in formula (14), ω1And ω2Respectively representing the weight of each sub-objective function in the total objective function, and comprehensively considering the weight value according to the actual situation. Definition of ω1And omega2The sum is 1. From equation (14), it can be derived that the larger the total objective function value f, the better the quality of the solution, and vice versa.
(2) Coding mode and genetic manipulation
The coverage optimization control of the wireless sensor network adopts a genetic algorithm as a prototype, so that the working state of the sensor node needs to be coded, and the coverage optimization control adopts a binary coding mode and is represented by a bit string, namely a ═ a-1,a2,...,ai,...,aN). When the sensor node si participates in the sensor network coverage, namely the node is in a working state, the ai code is 1, otherwise, the ai code is 0.
One individual in the population p ═ a1,a2,...,ai,...,aN) I.e. represents a coverage control selection scheme. For example, randomly deploying 9 sensors in a monitored areaThe node, only s2, s3, s6, s8, s9 are in working state, and can be represented as: 011001011, as shown in fig. 4.
After the individual is encoded, the following operations are taken:
1) selecting operation: the selection aims to select individuals with high fitness as much as possible and reduce the population evolution time. The method selects individuals with high fitness by adopting a roulette mode. The probability that individual i is selected is:
in the formula (15), fiRepresenting fitness value of individual i; n denotes the population size.
When in selection, individuals are selected in a random equidistant mode, and the position of a first pointer is a uniform random number in a [0,1/N ] interval.
2) Crossover and mutation operations
The crossover operation is to exchange and combine partial genes of two chromosomes of a parent, and to transmit excellent genes to offspring to form new excellent individuals, and the crossover operation is a main method for searching new individuals. Reasonable crossover operators can accelerate the population search speed. The single point crossing method is used herein, and the crossing points are randomly generated.
And performing mutation operation after the cross operation is finished, wherein the mutation operation forms a new individual by changing a certain gene or certain genes of a certain individual, and forces an algorithm to develop a new search area to avoid premature convergence. In this chapter, mutation into its allele is used for mutation, and only one gene is mutated, and if the gene before mutation is 0, the gene after mutation is 1.
(3) End conditions
The condition for terminating iteration of the basic genetic algorithm is maximum evolution algebra, and if the maximum evolution algebra is not selected properly, the evolution stops under the condition that the evolution is not converged or redundant genetic operation is carried out under the condition that the evolution is converged. Therefore, the multi-population genetic algorithm adopts the method of keeping the optimal individuals in the algebra to be unchanged to the end iteration condition at least, the appearance of the optimal individuals represents that the evolution is finished, and redundant operation is not needed, so that the operation time is reduced, and the occurrence of an immature phenomenon is avoided.
Specifically, the energy consumption of the sensor nodes is directly influenced by the difference of the networking modes of the wireless sensors. The high-efficiency routing algorithm can reduce the energy consumption of nodes in the wireless sensor network, and the low-power-consumption self-adaptive clustering hierarchical LEACH, the clustering HEED with fixed cluster radius and other protocols in the clustering routing algorithm can prolong the service life of the sensor network from different angles. LEACH requires that energy initially stored by each node is equal during networking, but cluster scale is random, energy obtained by the nodes through an energy acquisition technology is different, the assumption that residual energy of each node is completely the same is not considered, and nodes with extremely low energy are selected as cluster heads, so that the nodes fail to operate, and the network cannot operate. According to the HEED, the cluster head is selected according to 2 parameters of the residual energy and the node density, the selection of the cluster head needs a large amount of calculation on the node, the power consumption of the node is increased, and the problem of how to realize remote data transmission between the cluster head and a base station is not considered. The static networking mode of the energy-saving wireless sensor replaces complex dynamic networking, so that the program of the node is simple, the energy consumption is reduced, and the sensing network can run persistently. In the network, the convergent nodes and the relay nodes are used for replacing cluster heads in the dynamic networking, so that the link of competing the cluster heads among the nodes is avoided. The division of the work of each type of nodes is different, and the working mode setting of the sensor nodes is different. And the working time of the sensor node is reduced, so that the power consumption of the sensor node is reduced.
The energy consumption of the sensor node mainly comes from: the device comprises a sensitive unit (sensor), a wireless communication module and a power management circuit. One duty cycle of a sensor node can be divided into: data collection, data transmission and dormancy stage. When each part of the wireless sensor is designed, the time parameter and the current parameter are reduced as much as possible, and the energy consumption can be reduced. The optimization of energy is realized by adjusting the working time of each part through programming.
The wireless sensor has the largest energy consumption during communication, and the data transmission quantity of the communication module is reduced as much as possible, so that the energy consumption is reduced. If multiple nodes transmit simultaneously, they overlap, resulting in a received signal that is illegible and requires retransmission to properly transmit the information to the destination. Retransmissions due to collisions result in wasted energy. When a plurality of nodes send data to the same sink node, the automatic retransmission time difference can be adjusted by the automatic retransmission time delay, and in the next data sending period, a time difference exists between the data of different nodes, so that the data collision is avoided. When the wireless communication module works in transmitting and receiving, the wireless communication module always keeps a data transmitting and receiving state, and the power consumption is large; when entering the sleep state, the transmitting and receiving functions are closed to reduce power consumption. Therefore, the timer is controlled to switch between the working state and the sleep 2 states according to needs, and low power consumption is achieved.
Specifically, the dust concentration prediction model optimizes the initial weight of the BP neural network based on an improved genetic algorithm, and then performs back propagation operation adjustment to obtain a generalized neural network model serving as the optimized dust concentration prediction model.
The genetic algorithm is a random optimization search method which is evolved by taking advantage of the evolution law (survival of fittest and high-and-low-rejection genetic mechanism) of the biological world. The basic algorithm flow comprises the operations of generating groups, selecting, crossing, mutating and the like. However, the traditional genetic algorithm also has the problems of slow convergence, easy precocity and the like. Aiming at the problem, a self-adaptive adjusting method of the cross probability and the mutation probability is provided. An adaptive variation probability function and adaptive cross probability functions Pc and Pm are introduced. It can be seen that the adaptive variation probability function and the cross probability function are a smooth curve on a plane, when the fitness of an individual is greater than the average fitness of a population, a smaller variation probability can be obtained by the adaptive variation and the cross probability function, such a design is beneficial to keeping excellent genes of the individual, and when the fitness of the individual is less than the average fitness of the population, a larger variation probability can be obtained by the adaptive variation probability function, such a design is beneficial to aggravating the variation of the poor individual. The adaptive mutation probability function and the adaptive cross probability function are as follows:
wherein: pcmax and Pmmax are the maximum cross probability and the maximum variation probability in the population respectively; f' is the fitness of the time; both β and η are adaptive coefficient factors that control the specific shape of the adaptive function. The degree of evolution is controlled by the variation probability function and the cross probability function, so that the defects of local optimization and prematurity of the genetic algorithm are overcome.
And optimizing the initial weight of the BP neural network by using an improved genetic algorithm, so that the method that the initial weight of the traditional BP neural network is randomly given is changed. The initial weight value optimized by the improved genetic algorithm is the solution after optimization in the space, then the optimal solution is substituted into the weight value of the neural network, and then the back propagation operation adjustment is carried out, so that a model with higher generalization capability than that of the traditional BP neural network can be obtained. The mathematically optimal solution model of the neural network model can be abstracted as:
wherein,
is the expected value of the neural network output.
The constraint conditions are as follows:
wherein: e is a loss function; y isk(xi) Actual values output for the neural network; ω and upsilon are vectors of m o and o n, respectively. The weight upsilon of the neural network can be obtained through an improved genetic algorithmjkAnd omegaijThe specific training steps are as follows:
(1) and setting the structure of the initial neural network, wherein the structure comprises the node number of an input layer, a hidden layer and an output layer.
(2) Setting parameters required for the genetic algorithm to proceed: the number N of the initial population, the final iteration algebra T, the initial cross rate Pc0 and the variation rate Pm0, the coefficients beta and eta of the adaptive cross function and the variation function, the length L of the chromosome and the like.
(3) Generating a primary generation population.
(4) And carrying out selective crossing and mutation operation on the population, wherein the mutation probability and the crossing probability are determined by using an adaptive function method of the previous subsection, namely when the fitness of the individual is less than the average fitness of the population, adaptively giving a larger crossing probability and a larger mutation probability, and conversely, adaptively giving a smaller crossing probability and a smaller mutation probability.
(5) If the genetic algebra is reached or is smaller than the target function value, the genetic algorithm is ended, and the obtained weight is input into the BP neural network. And (4) if the condition is not met, re-executing the step (4).
(6) And substituting the optimized weight into a BP neural network to perform back propagation operation, and adjusting the weight. And finally obtaining a generalized neural network model.
Specifically, the dust concentration prediction model is further optimized through dust concentration information and environmental information.
Example 2: as shown in fig. 5, the application also discloses a thermal power plant dust real-time monitoring method based on energy-saving optimization,
s1: collecting dust concentration information and environmental information of a thermal power plant;
specifically, the method for obtaining the dust concentration includes: and continuously detecting the dust by using a light transmission method to obtain the dust concentration.
Specifically, data fusion and optimization processing are performed on the dust concentration information and the environment information by constructing a support function.
The data layer fusion is the preprocessing of the data of the same sensor, is the fusion process of the lowest level and the lowest layer in the multilevel data fusion, and mainly solves the problems of false alarm, missing detection, redundancy, large calculated amount and the like possibly occurring in the original data. Particularly for the monitoring data of the wireless sensor network, the negative influence of the complex environment of high-concentration dust and the factors of the sensor on the decision result can be effectively solved through data layer fusion, and high-quality data resources and more credible decision basis are provided for the decision layer fusion.
The data layer fusion algorithm mainly comprises a weighted average algorithm, a Bayes algorithm, a Kalman filtering method and the like, which are classical data processing algorithms. However, two problems are exposed in practical application, one is that most algorithms presuppose data to be subjected to normal distribution, but a large number of experiments show that most of instrument measurement data do not conform to the normal distribution, but are between uniform and normal; secondly, the algorithm may not be able to cull individual outliers. In order to solve the problems, a fusion algorithm based on a support matrix is provided, which only utilizes the information contained in the data, avoids the assumption of normal distribution and can accurately fuse the data.
The concept of data support is introduced first. For any set of measured data X1,X2,...,XnIn particular, XiTo XjIs from XiFrom the angle of (1), XjIs the likely degree of the real data. Under the concept of support, the trueness of data is determined by itself, i.e. XiThe higher the authenticity of (a), the higher the degree of support by the rest of the data. On the basis, the relative distance between data is defined:
dij＝|Xi-Xj|,(i,j＝1,2,...,n) (20)
and then defining a support function:
in the formula, max { d {ijIs the maximum value of the relative distance between the data. As can be seen from the formula (16), the support qijWith relative distance dijIs increased and decreased. This indicates the difference between the dataThe more obvious the difference is, the smaller the mutual support degree is, and the obvious principle is met. For the data fusion problem, a support degree matrix is established
The support matrix Q is used for obtaining the comprehensive support degree of the ith data by other data, namely determining the weight coefficient Wi of the ith data in the whole data. As can be seen from the information sharing principle,
since Wi should synthesize qi1,qi2,...,qinThe general information of (2) is known from probability source combination theory that a group of nonnegative real numbers r must existi1,ri2,...,rinSo that
Wi＝r1qi1+r2qi2+...+rnqin (23)
Let W ═ W1W 2 … Wn ] T, R ═ R1R2 … Rn ] T
Equation (19) is then expressed in matrix form as: w ═ QR
Since the support matrix Q is a non-negative matrix, it must have a non-negative maximum eigenvalue λmaxAnd is formed bymaxMaximum eigenvalue λ can be obtained when R is QRmaxCorresponding feature vector R ═ R1R2 … Rn]T, the weight coefficient Wi of the ith data Xi is:
then X1,X2,...,XnThe final fusion result for this set of data is:
X＝W1X1+W2X2+...+WnXn
s2: establishing a dust concentration prediction model based on a BP neural network, and optimizing the dust concentration prediction model based on an improved genetic algorithm;
s3: and carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model.
The above-described embodiments are merely illustrative of the preferred embodiments of the present application, and do not limit the scope of the present application, and various modifications and improvements made to the technical solutions of the present application by those skilled in the art without departing from the spirit of the present application should fall within the protection scope defined by the claims of the present application.
Claims (10)
1. The utility model provides a dust real-time monitoring system of thermal power plant based on energy-conserving optimization which characterized in that includes: the system comprises a data acquisition module, a model construction module and a monitoring module;
the data acquisition module is used for acquiring dust concentration information and environmental information of the thermal power plant;
the model construction module is used for establishing a dust concentration prediction model based on a BP neural network and optimizing the dust concentration prediction model based on an improved genetic algorithm;
the monitoring module is used for carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model;
the data acquisition module, the model building module and the monitoring module are in data transmission in a wireless connection mode.
2. The energy-saving optimization-based real-time monitoring system for dust of a thermal power plant according to claim 1, wherein the data acquisition module comprises a detection unit and a sensor unit;
the detection unit is used for detecting the dust concentration by a light transmission method;
the sensor unit is used for receiving the dust concentration, collecting the environmental information and sending the dust concentration and the environmental information to the monitoring module;
the detection unit and the sensor unit are in wireless connection for data transmission.
3. The energy-saving optimization-based real-time monitoring system for dust of a thermal power plant according to claim 2, wherein the sensor unit performs data fusion and optimization processing on the dust concentration information and the environmental information by using a support function.
4. The energy-saving optimization-based real-time monitoring system for dust of a thermal power plant according to claim 2, wherein the detection unit comprises: the device comprises a laser, a photoelectric detector, a signal processing circuit, a singlechip, a computer, a power supply and an air blowing device;
the laser is used for emitting laser beams;
the photoelectric detector is used for receiving the laser beam in the detected dust area and converting the optical signal of the laser beam into an electric signal;
the air blowing device is used for cleaning a window of the laser and a window of the photoelectric detector;
the signal processing circuit is used for receiving the electric signal and converting the electric signal into a digital signal;
the single chip microcomputer is used for preprocessing the digital signal to obtain the dust concentration data and sending the dust concentration data to the computer;
the computer displays the received dust concentration data by using LabVIEW software.
5. The energy-saving optimization-based real-time thermal power plant dust monitoring system according to claim 4, wherein an auxiliary detector is arranged to receive ambient light when the dust concentration is detected.
6. The energy-saving optimization-based real-time thermal power plant dust monitoring system according to claim 1, wherein the optimization method of the dust concentration prediction model comprises the following steps: optimizing the initial weight of the BP neural network based on an improved genetic algorithm, and then performing back propagation operation adjustment to obtain a generalized neural network model serving as the optimized dust concentration prediction model; the improved genetic algorithm is an improved self-adaptive adjusting method based on cross probability and mutation probability.
7. The energy-saving optimization-based real-time thermal power plant dust monitoring system according to claim 1, wherein the dust concentration prediction model is further optimized through the dust concentration information and the environmental information.
8. A thermal power plant dust real-time monitoring method based on energy-saving optimization is characterized in that,
collecting dust concentration information and environmental information of a thermal power plant;
establishing a dust concentration prediction model based on a BP neural network, and optimizing the dust concentration prediction model based on an improved genetic algorithm;
and carrying out real-time prediction and early warning on the dust concentration information and the environmental information based on the dust concentration prediction model.
9. The energy-saving optimization-based real-time monitoring method for thermal power plant dust according to claim 8, wherein the method for acquiring the dust concentration comprises the following steps: and continuously detecting the dust by using a light transmission method to obtain the dust concentration.
10. The energy-saving optimization-based real-time monitoring method for thermal power plant dust according to claim 8, wherein the dust concentration information and the environmental information are subjected to data fusion and optimization processing by constructing a support function.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111463373.9A CN114167726A (en) | 2021-12-03 | 2021-12-03 | Thermal power plant dust real-time monitoring system and method based on energy conservation optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111463373.9A CN114167726A (en) | 2021-12-03 | 2021-12-03 | Thermal power plant dust real-time monitoring system and method based on energy conservation optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114167726A true CN114167726A (en) | 2022-03-11 |
Family
ID=80482576
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111463373.9A Pending CN114167726A (en) | 2021-12-03 | 2021-12-03 | Thermal power plant dust real-time monitoring system and method based on energy conservation optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114167726A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115310361A (en) * | 2022-08-16 | 2022-11-08 | 中国矿业大学 | Method and system for predicting underground dust concentration of coal mine based on WGAN-CNN |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202793980U (en) * | 2012-08-27 | 2013-03-13 | 陕西理工学院 | Distributed wireless dust measuring instrument |
| CN104345020A (en) * | 2014-10-24 | 2015-02-11 | 海安能度软件科技有限公司 | Dust detection alarm system |
| CN107358363A (en) * | 2017-07-17 | 2017-11-17 | 哈尔滨理工大学 | Coal work incidence of disease Forecasting Methodology based on radial basis function neural network built-up pattern |
| CN107688208A (en) * | 2017-09-08 | 2018-02-13 | 重庆三峡学院 | A kind of environmental monitoring dust granules system |
| CN111474094A (en) * | 2020-03-20 | 2020-07-31 | 淮阴工学院 | Intelligent dust concentration detection system |
| CN113419425A (en) * | 2021-07-07 | 2021-09-21 | 哈尔滨理工大学 | Intelligent regulation and control method for high-speed electric spindle water cooling system |
| CN113643515A (en) * | 2021-07-30 | 2021-11-12 | 张家港智能电力研究院有限公司 | Combustible dust automatic identification judges and alarm system |
-
2021
- 2021-12-03 CN CN202111463373.9A patent/CN114167726A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN202793980U (en) * | 2012-08-27 | 2013-03-13 | 陕西理工学院 | Distributed wireless dust measuring instrument |
| CN104345020A (en) * | 2014-10-24 | 2015-02-11 | 海安能度软件科技有限公司 | Dust detection alarm system |
| CN107358363A (en) * | 2017-07-17 | 2017-11-17 | 哈尔滨理工大学 | Coal work incidence of disease Forecasting Methodology based on radial basis function neural network built-up pattern |
| CN107688208A (en) * | 2017-09-08 | 2018-02-13 | 重庆三峡学院 | A kind of environmental monitoring dust granules system |
| CN111474094A (en) * | 2020-03-20 | 2020-07-31 | 淮阴工学院 | Intelligent dust concentration detection system |
| CN113419425A (en) * | 2021-07-07 | 2021-09-21 | 哈尔滨理工大学 | Intelligent regulation and control method for high-speed electric spindle water cooling system |
| CN113643515A (en) * | 2021-07-30 | 2021-11-12 | 张家港智能电力研究院有限公司 | Combustible dust automatic identification judges and alarm system |
Non-Patent Citations (8)
| Title |
|---|
| 刘晓悦等: "基于云计算的煤矿瓦斯浓度短期预测", 《控制工程》 * |
| 夏靖波等: "基于遗传算法神经网络流量测量", 《东北大学学报(自然科学版)》 * |
| 姜倩倩等: "一种节能无线传感器网络", 《太赫兹科学与电子信息学报》 * |
| 王文韬: "改进的高速铁路安全状态协同监测数据处理方法研究", 《计算机测量与控制》 * |
| 白鹏，等: "改进遗传算法优化的BP神经网络高炉煤气预测", 《机械工程与自动化》 * |
| 祁育仙等: "混合WSNs中基于多目标优化的覆盖控制算法", 《传感器与微系统》 * |
| 赵广元等: "粒子群算法优化BP神经网络的粉尘浓度预测", 《测控技术》 * |
| 赵欣然等: "大量程在线式粉尘浓度监测系统设计", 《中国测试》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115310361A (en) * | 2022-08-16 | 2022-11-08 | 中国矿业大学 | Method and system for predicting underground dust concentration of coal mine based on WGAN-CNN |
| CN115310361B (en) * | 2022-08-16 | 2023-09-15 | 中国矿业大学 | Underground coal mine dust concentration prediction method and system based on WGAN-CNN |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2021174971A1 (en) | Water environment risk prediction and early warning method | |
| Wang et al. | Research on prediction of environmental aerosol and PM2. 5 based on artificial neural network | |
| Kang et al. | Application of BP neural network optimized by genetic simulated annealing algorithm to prediction of air quality index in Lanzhou | |
| CN106202946A (en) | Clinker free calcium levels Forecasting Methodology based on degree of depth belief network model | |
| CN101867960A (en) | Comprehensive evaluation method for wireless sensor network performance | |
| CN111027256A (en) | Bridge risk prediction method and system based on vehicle load spatial distribution | |
| CN110322014A (en) | A kind of finished cement specific surface area prediction technique based on BP neural network | |
| CN114167726A (en) | Thermal power plant dust real-time monitoring system and method based on energy conservation optimization | |
| CN113094988A (en) | Data-driven slurry circulating pump operation optimization method and system | |
| CN118095865B (en) | AI large model driven emergency fire-fighting fire risk assessment and early warning method | |
| CN108760592A (en) | A kind of unburned carbon in flue dust On-line Measuring Method based on BP neural network | |
| CN113408659A (en) | Building energy consumption integrated analysis method based on data mining | |
| CN112990546A (en) | Chemical plant power transformer fault prediction method based on particle swarm and neural network | |
| CN112016249A (en) | SCR denitration system bad data identification method based on optimized BP neural network | |
| CN104599026A (en) | Supplier selection method | |
| CN109376903A (en) | A kind of PM2.5 concentration value prediction technique based on game neural network | |
| CN117350146A (en) | GA-BP neural network-based drainage pipe network health evaluation method | |
| Cao et al. | Field detection of indoor fire threat situation based on LSTM-Kriging network | |
| Sung et al. | IHPG algorithm for efficient information fusion in multi-sensor network via smoothing parameter optimization | |
| CN116771425A (en) | Detection method and system for spatial distribution of concentration of gas in coal mine tunnel | |
| Wen et al. | Design of an intelligent alarm system based on multi-sensor data fusion | |
| Reddy et al. | A smart environment monitoring framework using big data and IoT | |
| CN113487114A (en) | GA-SVM early warning model system based on NB-IoT industrial ecological environment monitoring | |
| Tao et al. | [Retracted] Optimization Analysis of Power Coal‐Blending Model and Its Control System Based on Intelligent Sensor Network and Genetic Algorithm | |
| CN103675012A (en) | Industrial melt index soft measurement instrument and method based on BP particle swarm optimization |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220311 |
|
| RJ01 | Rejection of invention patent application after publication |