CN111880408A - Sludge drying chamber control method and system based on multi-sensor data fusion - Google Patents
Sludge drying chamber control method and system based on multi-sensor data fusion Download PDFInfo
- Publication number
- CN111880408A CN111880408A CN202010758249.4A CN202010758249A CN111880408A CN 111880408 A CN111880408 A CN 111880408A CN 202010758249 A CN202010758249 A CN 202010758249A CN 111880408 A CN111880408 A CN 111880408A
- Authority
- CN
- China
- Prior art keywords
- drying
- drying chamber
- sensor
- value
- parameters
- 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
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
-
- C—CHEMISTRY; METALLURGY
- C02—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
- C02F—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
- C02F11/00—Treatment of sludge; Devices therefor
- C02F11/12—Treatment of sludge; Devices therefor by de-watering, drying or thickening
- C02F11/13—Treatment of sludge; Devices therefor by de-watering, drying or thickening by heating
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Medical Informatics (AREA)
- Automation & Control Theory (AREA)
- Mechanical Engineering (AREA)
- Hydrology & Water Resources (AREA)
- Environmental & Geological Engineering (AREA)
- Water Supply & Treatment (AREA)
- Chemical & Material Sciences (AREA)
- Organic Chemistry (AREA)
- Drying Of Solid Materials (AREA)
Abstract
The invention relates to the technical field of Internet of things, in particular to a sludge drying chamber control method and system based on multi-sensor data fusion. The method comprises the following steps: s1, collecting drying parameters of the drying chamber and running state parameters of the drying equipment in real time; s2, rejecting abnormal data in the parameters based on the thought of group support degree, and performing missing value processing on the pretreatment drying parameters and the running state parameters of the pretreatment drying equipment to obtain the pretreatment drying parameters and the running state parameters of the pretreatment drying equipment; s3, constructing a data sample; s4, outputting the data sample to a drying chamber model of a pre-established multi-sensor data fusion algorithm, and outputting judgment parameters of a drying effect; s5, controlling the drying device of the drying chamber according to the judgment parameter of the drying effect. The method can improve the accuracy of prediction and the real-time performance of control, and can operate at optimal efficiency on the premise of meeting the set drying target, thereby reducing the energy consumption to the maximum extent.
Description
Technical Field
The invention relates to the technical field of Internet of things, in particular to a sludge drying chamber control method and system based on multi-sensor data fusion.
Background
With the emergence and development of emerging technologies, the automation level of industrial machinery is continuously improved, so that the amount of generated information is increased, the data to be processed becomes huge, and therefore, the intelligent control system also needs to be continuously optimized, for example, the corresponding control method and the algorithm need to be continuously improved to meet the current requirements. In the technical field of control, China starts late, and a plurality of technologies need to be continuously developed and perfected. For example, a traditional PID algorithm is mostly used in an automatic control system, and although the algorithm is classical, in some practical applications, due to the complexity of a research object and time-varying uncertainty, it is difficult to accurately establish a mathematical model, and efficient control requirements cannot be met.
Along with the continuous development of social economy, the natural ecological environment is inevitably polluted by industrial production, and the shortage of more energy resources is caused by the large consumption of the energy resources, and even the threat of exhaustion is faced. In modern industrial development, environmental protection is particularly important, so that low energy consumption and environmental protection are the right development directions in the future development of the industrial industry. For example, waste materials generated in industrial production are treated, and no matter the waste materials are recycled or subjected to harmless treatment, environmental protection and low energy consumption are achieved as much as possible.
Disclosure of Invention
The invention aims to solve the problem that the traditional PID algorithm cannot adjust a control strategy according to the real-time change of a research object to realize accurate control, and provides a sludge drying chamber control method and system based on multi-sensor data fusion based on data fusion and a BP neural network.
In order to achieve the above purpose, the invention provides the following technical scheme:
a sludge drying chamber control method based on multi-sensor data fusion comprises the following steps:
s1, acquiring drying parameters of the drying chamber in real time by adopting a plurality of different types of sensors, and acquiring running state parameters of the drying equipment;
s2, rejecting abnormal data in the drying parameters and the drying equipment running state parameters based on the group support degree idea, and performing missing value processing on the drying parameters and the drying equipment running state parameters to obtain pre-processing drying parameters and pre-processing drying equipment running state parameters;
s3, constructing a data sample based on the pretreatment drying parameters and the operation state parameters of the pretreatment drying equipment;
s4, outputting the data sample to a drying chamber model of a pre-established multi-sensor data fusion algorithm, and outputting a judgment parameter of a drying chamber material drying effect;
s5, controlling the drying device of the drying chamber according to the judgment parameter of the drying effect of the material in the drying chamber.
Further, the drying parameters of the drying chamber comprise drying chamber inlet air quantity, drying chamber outlet air temperature, wet material water content, drying chamber drying temperature, drying chamber relative humidity, water removal quantity and dry material water content; the drying equipment running state parameters comprise working voltage and power, and each drying parameter is acquired by a plurality of homogeneous sensors.
Further, in step S2, the step of rejecting abnormal data in the drying parameter and the drying apparatus operating state parameter based on the idea of population support degree includes:
the method comprises the following steps that firstly, a plurality of measured values collected by the same sensor are obtained within a period of time, the mean value of the measured values is calculated, and the measured value closest to the mean value is used as the value with the highest effectiveness;
and secondly, calculating the difference value between each measured value and the value with the highest effectiveness, and rejecting the corresponding measured value when the difference value is greater than a support degree threshold value and does not meet the group support degree.
As a preferred embodiment of the present invention, in step S2, the step of performing missing value processing on the pretreatment drying parameter and the pretreatment drying apparatus operating state parameter includes: and in a preset time, obtaining the average value of the data acquired by each sensor, and supplementing the missing value in the data acquired by the corresponding sensor by using each average value of the data.
As a preferred scheme of the present invention, the drying chamber model of the pre-established multi-sensor data fusion algorithm includes two parts: a first level of fusion and a second level of fusion,
the first stage of fusion includes:
processing the same type of parameters collected by a plurality of homogeneous sensors in the data sample by using a self-adaptive weighted average algorithm to obtain a fusion value of each homogeneous sensor parameter;
the second stage of fusion comprises the following steps:
and combining the fusion values of the homogeneous sensor parameters into an array, and inputting the array into a BP neural network to obtain a judgment parameter of the drying effect of the material in the drying chamber.
As a preferred embodiment of the present invention, the BP neural network includes an input layer, a hidden layer and an output layer, where the number of nodes of the hidden layer is n,
x is the number of input layer nodes, m is the number of output layer nodes, a is a constant between 1 and 10; the excitation function of the neuron is a sigmoid function, and after the learning rate and the training times are determined, the value output by the output layer is a judgment parameter of the drying effect of the material in the drying chamber.
As a preferred embodiment of the present invention, the specific calculation steps of each type of fusion value are:
s41, listing an expression of the total mean square error of the homogeneous sensors according to the mean value, the measured value and the weight value of the data collected by each homogeneous sensor;
s42, adopting a multivariate function extremum solving theory to calculate the corresponding weight value of each sensor in the homogenous sensor when the total mean square error is minimum;
and S43, calculating the fusion value of each homogeneous sensor parameter by adopting an adaptive weighted average algorithm.
As a preferred embodiment of the present invention, the expression of the total mean square error of the homogeneous sensor is:
wherein σ2Is the total mean square error, W, of the homogeneous sensor1,W2,…,WnIs the corresponding weight value of each sensor in the homogeneous sensor; sigma1 2,σ2 2,…σn 2The variance of data collected by each sensor in the homogeneous sensor; n is the number of sensors in a homogenous sensor.
Based on the same conception, the invention also provides a sludge drying chamber control system based on multi-sensor data fusion, which comprises at least one processor and a memory which is in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform any of the methods described above.
Compared with the prior art, the invention has the beneficial effects that:
1. the method comprises the steps of collecting drying parameters of a drying chamber through a plurality of homogeneous sensors, collecting working parameters of drying chamber equipment, rejecting abnormal data in the drying parameters and the drying equipment running state parameters based on a group support degree thought, and performing missing value processing on the drying parameters and the drying equipment running state parameters to construct a data sample; outputting the data sample to a drying chamber model of a pre-established multi-sensor data fusion algorithm, and outputting a judgment parameter of the drying effect of the drying chamber material; and controlling the drying equipment of the drying chamber according to the judgment parameter of the drying effect of the material in the drying chamber. The drying chamber model of the multi-sensor data fusion algorithm is a two-stage fusion model, the accuracy of prediction and the real-time performance of control can be improved through the processing of the method, and the drying chamber model can operate at the optimal efficiency on the premise of meeting related set drying targets, so that the energy consumption is reduced to the maximum extent.
2. According to the method, data of the homogeneous sensor are fused by adopting a self-adaptive weighted average algorithm, the fused parameters are combined into an array and input into a BP neural network, so that a judgment parameter of a drying effect is obtained, and drying equipment of a drying chamber is controlled according to the judgment parameter of the drying effect. The method adopts a self-adaptive weighted average algorithm to fuse and balance the difference of data acquired by different sensors in a homogeneous sensor, sets the sensor with accurate acquired data as a higher weight value, sets the sensor with larger acquired data deviation as a lower weight value, so that the data more accurately reflects the state of a drying chamber, obtains a judgment parameter of a drying effect through a BP neural network, and ensures the operating efficiency because the network parameter is an optimal parameter trained by a large number of samples, the output result is adaptively adjusted according to the operating state of the system, and the operating parameter of equipment is adjusted according to a predicted value.
Description of the drawings:
FIG. 1 is a flow chart of a method for controlling a sludge drying chamber based on multi-sensor data fusion in embodiment 1 of the present invention;
FIG. 2 is a schematic structural diagram of a two-stage fusion model in embodiment 1 of the present invention;
fig. 3 is a table showing correspondence between spatial parameter values of samples and drying effects in embodiment 1 of the present invention.
Detailed Description
The present invention will be described in further detail with reference to test examples and specific embodiments. It should be understood that the scope of the above-described subject matter is not limited to the following examples, and any techniques implemented based on the disclosure of the present invention are within the scope of the present invention.
Example 1
A flow chart of a sludge drying chamber control method based on multi-sensor data fusion is shown in figure 1, and comprises the following steps:
s1, collecting drying parameters of the drying chamber and running state parameters of the drying equipment in real time. The drying parameters of the drying chamber comprise the inlet air quantity of the drying chamber, the water content of the material entering the drying chamber, the water content of the material at the outlet of the drying chamber, the temperature and humidity at the air outlet of the drying chamber, the water removal quantity and the like. The operation state parameters of the drying equipment comprise working voltage, power, drying chamber set temperature, drying operation time and the like, each drying parameter is acquired by a plurality of sensors, for example, the temperature of an air inlet is not measured by one temperature sensor, but a plurality of temperature sensors are simultaneously used for detecting the temperature of the air inlet, a plurality of corresponding humidity sensors are simultaneously used for detecting the humidity of the air inlet, a plurality of air speed sensors are simultaneously used for detecting the air volume of the drying chamber and the like, and therefore the acquired data volume is large. In addition, the sensors of the same type are named as homogeneous sensors, for example, a plurality of temperature sensors are temperature homogeneous sensors, a plurality of humidity sensors are humidity homogeneous sensors, and therefore accurate description is conveniently carried out during the subsequent fusion processing of the sensor data of the same type.
S2, rejecting abnormal data in the drying parameters and the drying equipment running state parameters based on the group support degree thought, and performing missing value processing on the pretreatment drying parameters and the pretreatment drying equipment running state parameters to obtain pretreatment drying parameters and pretreatment drying equipment running state parameters. In the data fusion process, invalid abnormal data can cause the final fusion result to be inaccurate, so the invalid abnormal data should be removed. In order to accurately judge the effectiveness of the abnormal data, the thought of group support degree is introduced, and the distance of the method for judging the effectiveness of the abnormal data by using the group support degree is as follows:
assuming two measured values ai and aj, when ai and aj are closer, the higher the support degree of the two values is, which satisfies the following:
sup(a,b)∈[0,1]；
sup(a,b)＝sup(b,a)；
if | a-b | < | x-y |, then sup (a, b) > sup (x, y).
In all the sensor node data, there is a support degree, and the support degree reflects the support of the adjacent nodes in the area on the validity of the abnormal data. Suppose there are n sensors for temperature in the whole drying chamber, wherein the measurement value of the ith is marked as aiThe j-th measurement is denoted as aj. If a isiThe validity of the measured value is high, the higher the support degree of the measured value of other nodes is proved, and a isiThe higher the likelihood of being valid data.
As an embodiment, the step of deleting the abnormal data with the group support degree includes:
the method comprises the steps of firstly, acquiring a plurality of measured values collected by the same sensor in a period of time, solving the mean value of a plurality of data, and taking the measured value closest to the mean value as the value with the highest effectiveness;
and secondly, calculating the difference value between each measured value and the value with the highest effectiveness, and when the difference value is greater than the support degree threshold value, not meeting the group support degree, and rejecting the corresponding measured value as abnormal data.
Preferably, in step S2, the step of performing missing value processing on the pre-treatment drying parameter and the pre-treatment drying apparatus operating state parameter includes: and in a preset time, obtaining the average value of the data acquired by each sensor, and supplementing the missing value in the data acquired by the corresponding sensor by using each average value of the data.
S3, constructing a data sample based on the pretreatment drying parameters and the operation state parameters of the pretreatment drying equipment. For example, an isomeric array X is formed by drying chamber inlet air volume, drying chamber outlet air temperature, wet material water content, drying chamber drying temperature, drying chamber relative humidity, dewatering amount, dry material water content, drying device operating voltage and drying device operating power, and is used for inputting a BP neural network and calculating judgment parameters of a drying effect, wherein X is { X1, X2, X3, X4, X5, X6, X7, X8 and X9 }.
And S4, outputting the heterogeneous array to a drying chamber model of a pre-established multi-sensor data fusion algorithm, and outputting judgment parameters of the drying effect of the drying chamber material. The drying chamber model of the multi-sensor data fusion algorithm is a two-stage fusion model and comprises a first-stage fusion model and a second-stage fusion model. The structural schematic diagram of the two-level fusion model is shown in fig. 2.
And S5, controlling the drying equipment of the drying chamber according to the predicted value of the material drying.
As a preferred scheme, the pre-established drying chamber model of the multi-sensor data fusion algorithm comprises a two-stage fusion treatment process,
the first stage of fusion includes: processing the same type parameters collected by a plurality of homogeneous sensors in a data sample by using a self-adaptive weighted average algorithm to obtain a fusion value of each homogeneous sensor;
the second stage of fusion comprises the following steps: and combining the fusion values of the homogeneous sensors into an array, and inputting the array into a BP neural network to obtain a judgment parameter of the drying effect of the material in the drying chamber.
The specific calculation steps of the fusion value of each homogeneous sensor are as follows:
s41, listing an expression of the total mean square error of the homogeneous sensor according to the mean value, the measured value and the weight value of the data collected by each sensor;
s42, adopting a multivariate function extremum solving theory to calculate the corresponding weight value of each sensor in the homogenous sensor when the total mean square error is minimum;
and S43, calculating the fusion value of the parameters of the plurality of sensors in the homogeneous sensor by adopting an adaptive weighted average algorithm.
Let the variance of n homogeneous sensors be σ1 2,σ2 2,…σn 2The weight of each sensor is W1,W2,…,WnMeasured values are respectively X1,X2,…,XnIndependent of each other and being an unbiased estimate of X, the fused values
Comprises the following steps:
and the weight W1,W2,…,WnIs 1, the expression is as follows:
the total mean square error sigma can be derived from the above conditions2Comprises the following steps:
to obtain the current total mean square error sigma2Each weight W at the minimumiAnd (3) solving an extreme value theory by a multivariate function:
at this time, the corresponding minimum mean square error σ2 minComprises the following steps:
therefore, the variance of each sensor can be obtained, and the corresponding weight value can be obtained and used in subsequent data fusion to obtain optimal fusion data.
The factor with the largest drying effect in the sample subset is the temperature and the humidity of the drying chamber, so that more importance is placed on the acquisition and monitoring of temperature and humidity data, a plurality of temperature sensors and a plurality of humidity sensors are adopted, the data volume of the acquired temperature and humidity data is large, and the calculated amount is large if the data received by the sensors are directly input into a neural network for training or testing. Therefore, preferably, the collected values of the plurality of temperature sensors are fused by using an adaptive weighted average algorithm to obtain a fused value of the temperature homogeneous sensor. The acquisition values of a plurality of humidity sensors are fused by using a self-adaptive weighted average algorithm to obtain a fusion value of the humidity homogeneous sensor, and the fusion value of the temperature homogeneous sensor, the fusion value of the humidity homogeneous sensor and the type parameter are combined into an array and input into a BP neural network for training or testing. And after the optimized BP neural network is obtained through training, testing the input array by using the optimized BP neural network, and outputting a drying effect parameter beta of the drying chamber. Preferably, when the drying effect parameter beta range of the drying chamber is more than or equal to 85% and less than or equal to 75%, the system is considered to run well; when the beta is more than or equal to 85% and less than 100%, the system is considered to be very good in operation, when the beta is more than 60% and less than or equal to 75%, the system is considered to be qualified in operation, and when the beta is less than 60%, the system is considered to be unqualified in operation, the parameters of the equipment need to be adjusted.
The process of training the BP neural network is as follows:
firstly, determining the number of hidden layers of the model, the number of input parameters and the number of output parameters. Number of nodes in hidden layer
Where x ═ 7 is the number of nodes in the input layer, m ═ 1 is the number of nodes in the output layer, and a is a constant between 1 and 10, and in order to reduce the amount of computation, the number of hidden layers is set to 1, the number of nodes is 9, and the number of hidden layers is set to n ═ 9. In addition, after the weight, the threshold value, the excitation function of the neuron, the learning rate and the training times among the initialization neurons are determined, parameters can be input for training.
The initialization weight is set by adopting a random number: the random function takes [0,1] multiplied by 2-1 to obtain weight data of [ -1,1], and the weight data obey normal distribution. The first layer of weights is denoted by W and takes 63 values randomly, and the second layer takes 9 values.
The excitation function is sigmoid function, and the expression is
And connecting the input parameter X and the hidden layer to obtain an output layer. The output layer prediction value is calculated as:
L1＝sigmoid(x,W1)
L2＝sigmoid(L1,W2)
x is a value input to the BP neural network, W1 is a weight between the input layer and the hidden layer, W2 is a weight between the output layer and the hidden layer, and the iteration number steps is set to 10000; the learning rate lr is 0.11.
And secondly, randomly selecting 70% from the sample data set as a training set, using the rest 30% as a test set, initializing an input layer of the model, and normalizing the training sample.
Wherein, the sample parameters of the input layer are: the air quantity at the inlet of the drying chamber, the water content of the material entering the drying chamber, the water content of the material at the outlet of the drying chamber, the temperature and humidity at the air outlet of the drying chamber, the water removal quantity and the working power of the drier.
And finally, obtaining a prediction error value of the BP neural network model, and adjusting the weight between the hidden layer and the input layer according to the prediction error. And then the prediction error is transmitted to the previous layer of network in a reverse direction, and the weights of the hidden layer and the input layer are adjusted. And judging whether the result reaches a preset training target or not, if not, continuing the training, if so, finishing the training and finishing the model training.
Wherein, the calculation formula of the prediction error value is as follows:
wherein E is a prediction error value, y is an acquisition value, t is a predicted value of an output, X is a value input to the BP neural network, W is a weight of the BP neural network, WX is a product of the input value and the weight, and f () is an excitation function. After obtaining the loss value, the weight error is deduced reversely:
ΔW＝-ηE'＝ηXT(t-y)f'(WX)＝ηXT
where η is the learning rate, y is the collected value, t is the predicted value of the output, WX is the product of the input value and the weight, f '() is the inverse function of the excitation function f (), E' ═ XT(t-y)f'(WX)。
The relationship between the acquisition parameters and the drying effect obtained in the BP neural network training process is shown in FIG. 3. As can be seen from the table in fig. 3: the inlet air quantity of the drying chamber, the outlet air temperature of the drying chamber, the moisture content of wet materials, the drying temperature of the drying chamber, the relative humidity of the drying chamber, the water removal quantity and the moisture content of dry materials respectively have respective threshold value intervals, and when a sample space is an original data set, the drying effect is good.
Input into the sample space: the drying chamber has higher drying temperature and better drying effect under the condition of unchanged other parameters due to the inlet air quantity of the drying chamber.
Input into the sample space: the moisture content of the wet material, the moisture content of the dry material and the relative humidity of the drying chamber are low, and the drying effect is better under the condition that other parameters are not changed.
Input into the sample space: meanwhile, the air quantity at the inlet of the drying chamber is met, and the drying temperature of the drying chamber is higher; and the drying effect is optimal under the two conditions of the moisture content of the wet material, the moisture content of the dry material and the low relative humidity rate of the drying chamber.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.
Claims (9)
1. A sludge drying chamber control method based on multi-sensor data fusion is characterized by comprising the following steps:
s1, acquiring drying parameters of the drying chamber in real time by adopting a plurality of different types of sensors, and acquiring running state parameters of the drying equipment;
s2, rejecting abnormal data in the drying parameters and the drying equipment running state parameters based on the group support degree idea, and performing missing value processing on the drying parameters and the drying equipment running state parameters to obtain pre-processing drying parameters and pre-processing drying equipment running state parameters;
s3, constructing a data sample based on the pretreatment drying parameters and the operation state parameters of the pretreatment drying equipment;
s4, outputting the data sample to a drying chamber model of a pre-established multi-sensor data fusion algorithm, and outputting a judgment parameter of a drying chamber material drying effect;
s5, controlling the drying device of the drying chamber according to the judgment parameter of the drying effect of the material in the drying chamber.
2. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 1, wherein the drying parameters of the drying chamber comprise drying chamber inlet air quantity, drying chamber outlet air temperature, wet material water content, drying chamber drying temperature, drying chamber relative humidity, water removal quantity and dry material water content; the drying equipment running state parameters comprise working voltage and power, and each drying parameter is acquired by a plurality of homogeneous sensors.
3. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 2, wherein in the step S2, the step of rejecting abnormal data in the drying parameters and the drying equipment operation state parameters based on the group support degree comprises:
the method comprises the following steps that firstly, a plurality of measured values collected by the same sensor are obtained within a period of time, the mean value of the measured values is calculated, and the measured value closest to the mean value is used as the value with the highest effectiveness;
and secondly, calculating the difference value between each measured value and the value with the highest effectiveness, and rejecting the corresponding measured value when the difference value is greater than a support degree threshold value and does not meet the group support degree.
4. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 3, wherein in the step S2, the step of performing missing value processing on the pre-treatment drying parameter and the pre-treatment drying device operation state parameter comprises: and in a preset time, obtaining the average value of the data acquired by each sensor, and supplementing the missing value in the data acquired by the corresponding sensor by using each average value of the data.
5. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 4, wherein the pre-established drying chamber model of the multi-sensor data fusion algorithm comprises two parts: a first level of fusion and a second level of fusion,
the first stage of fusion includes:
processing the same type of parameters collected by a plurality of homogeneous sensors in the data sample by using a self-adaptive weighted average algorithm to obtain a fusion value of each homogeneous sensor parameter;
the second stage of fusion comprises the following steps:
and combining the fusion values of the homogeneous sensor parameters into an array, and inputting the array into a BP neural network to obtain a judgment parameter of the drying effect of the material in the drying chamber.
6. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 5, wherein the BP neural network comprises an input layer, a hidden layer and an output layer, wherein the number of nodes of the hidden layer is n,
x is the number of input layer nodes, m is the number of output layer nodes, a is a constant between 1 and 10; the excitation function of the neuron is a sigmoid function, and after the learning rate and the training times are determined, the value output by the output layer is a judgment parameter of the drying effect of the material in the drying chamber.
7. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 5, wherein the specific calculation steps of the fusion values of each type are as follows:
s41, listing an expression of the total mean square error of the homogeneous sensors according to the mean value, the measured value and the weight value of the data collected by each homogeneous sensor;
s42, adopting a multivariate function extremum solving theory to calculate the corresponding weight value of each sensor in the homogenous sensor when the total mean square error is minimum;
and S43, calculating the fusion value of each homogeneous sensor parameter by adopting an adaptive weighted average algorithm.
8. The method for controlling the sludge drying chamber based on the multi-sensor data fusion as claimed in claim 7, wherein the expression of the total mean square error of the homogeneous sensor is as follows:
wherein σ2Is homogeneousTotal mean square error of sensor, W1,W2,…,WnIs the corresponding weight value of each sensor in the homogeneous sensor; sigma1 2,σ2 2,…σn 2The variance of data collected by each sensor in the homogeneous sensor; n is the number of sensors in a homogenous sensor.
9. The sludge drying chamber control system based on multi-sensor data fusion is characterized by comprising at least one processor and a memory which is in communication connection with the at least one processor; the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010758249.4A CN111880408A (en) | 2020-07-31 | 2020-07-31 | Sludge drying chamber control method and system based on multi-sensor data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010758249.4A CN111880408A (en) | 2020-07-31 | 2020-07-31 | Sludge drying chamber control method and system based on multi-sensor data fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111880408A true CN111880408A (en) | 2020-11-03 |
Family
ID=73205788
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010758249.4A Pending CN111880408A (en) | 2020-07-31 | 2020-07-31 | Sludge drying chamber control method and system based on multi-sensor data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111880408A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112624560A (en) * | 2020-11-30 | 2021-04-09 | 广州晟启能源设备有限公司 | Automatic online detection and regulation method for sludge moisture content and drying system |
| CN112949905A (en) * | 2021-01-28 | 2021-06-11 | 国网浙江省电力有限公司信息通信分公司 | Method and device for predicting data of edge domain sensor |
| CN113742472A (en) * | 2021-09-15 | 2021-12-03 | 达而观科技(北京)有限公司 | Data mining method and device based on customer service marketing scene |
| CN113848705A (en) * | 2021-09-02 | 2021-12-28 | 上海建工集团股份有限公司 | Concrete mixture workability informatization intelligent regulation and control method |
| CN114035629A (en) * | 2021-10-20 | 2022-02-11 | 科大讯飞股份有限公司 | Control method of drying equipment and related equipment |
| CN117105502A (en) * | 2023-10-23 | 2023-11-24 | 威海华友节能科技有限公司 | Self-adaptive sludge treatment control system and control method based on machine learning |
| CN117930784A (en) * | 2024-03-18 | 2024-04-26 | 中城院(北京)环境科技股份有限公司 | Automatic control system and method for organic solid waste comprehensive treatment equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103992019A (en) * | 2014-06-04 | 2014-08-20 | 宋雷震 | System for drying sludge by virtue of diversified heat energy |
| CN104529130A (en) * | 2014-12-31 | 2015-04-22 | 北京沃特林克环境工程有限公司 | Solar sludge drying device |
| CN109634245A (en) * | 2018-12-29 | 2019-04-16 | 河南省福可思环保科技有限公司 | Unattended artificial intelligent control system for biochemical sludge dehydration and drying technique |
| CN111578662A (en) * | 2020-04-26 | 2020-08-25 | 上海净泥新能源科技有限公司 | Intelligent recognition type condensation dehumidification mummification system |
-
2020
- 2020-07-31 CN CN202010758249.4A patent/CN111880408A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103992019A (en) * | 2014-06-04 | 2014-08-20 | 宋雷震 | System for drying sludge by virtue of diversified heat energy |
| CN104529130A (en) * | 2014-12-31 | 2015-04-22 | 北京沃特林克环境工程有限公司 | Solar sludge drying device |
| CN109634245A (en) * | 2018-12-29 | 2019-04-16 | 河南省福可思环保科技有限公司 | Unattended artificial intelligent control system for biochemical sludge dehydration and drying technique |
| CN111578662A (en) * | 2020-04-26 | 2020-08-25 | 上海净泥新能源科技有限公司 | Intelligent recognition type condensation dehumidification mummification system |
Non-Patent Citations (4)
| Title |
|---|
| 刘君华编著: "《智能传感器系统》", 31 May 2010 * |
| 谭德坤 等: "《基于异常数据驱动的WSN簇内数据融合方法》", 《传感技术学报》 * |
| 赵妍著: "《面向大数据的挖掘方法研究》", 31 July 2016 * |
| 金溢: "《基于多源数据融合的草原环境监测模型及算法研究》", 《中国优秀硕士学位论文全文数据库 农业科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112624560A (en) * | 2020-11-30 | 2021-04-09 | 广州晟启能源设备有限公司 | Automatic online detection and regulation method for sludge moisture content and drying system |
| CN112624560B (en) * | 2020-11-30 | 2022-04-26 | 广州晟启能源设备有限公司 | Automatic online detection and regulation method for sludge moisture content and drying system |
| CN112949905A (en) * | 2021-01-28 | 2021-06-11 | 国网浙江省电力有限公司信息通信分公司 | Method and device for predicting data of edge domain sensor |
| CN113848705A (en) * | 2021-09-02 | 2021-12-28 | 上海建工集团股份有限公司 | Concrete mixture workability informatization intelligent regulation and control method |
| CN113848705B (en) * | 2021-09-02 | 2023-05-30 | 上海建工集团股份有限公司 | Intelligent informatization regulation and control method for workability of concrete mixture |
| CN113742472A (en) * | 2021-09-15 | 2021-12-03 | 达而观科技(北京)有限公司 | Data mining method and device based on customer service marketing scene |
| CN113742472B (en) * | 2021-09-15 | 2022-05-27 | 达而观科技(北京)有限公司 | Data mining method and device based on customer service marketing scene |
| CN114035629A (en) * | 2021-10-20 | 2022-02-11 | 科大讯飞股份有限公司 | Control method of drying equipment and related equipment |
| CN114035629B (en) * | 2021-10-20 | 2023-04-07 | 科大讯飞股份有限公司 | Drying equipment control method and related equipment |
| CN117105502A (en) * | 2023-10-23 | 2023-11-24 | 威海华友节能科技有限公司 | Self-adaptive sludge treatment control system and control method based on machine learning |
| CN117930784A (en) * | 2024-03-18 | 2024-04-26 | 中城院(北京)环境科技股份有限公司 | Automatic control system and method for organic solid waste comprehensive treatment equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111880408A (en) | Sludge drying chamber control method and system based on multi-sensor data fusion | |
| CN108898215B (en) | Intelligent sludge bulking identification method based on two-type fuzzy neural network | |
| CN108280552B (en) | Power load prediction method and system based on deep learning and storage medium | |
| CN111982302A (en) | Temperature measurement method with noise filtering and environment temperature compensation | |
| Peng et al. | A new Jacobian matrix for optimal learning of single-layer neural networks | |
| Abba et al. | Modeling of water treatment plant performance using artificial neural network: case study Tamburawa Kano-Nigeria | |
| Pham et al. | Modelling of Bunus regional sewage treatment plant using machine learning approaches | |
| Bloch et al. | Neural networks for process control and optimization: Two industrial applications | |
| CN116227689A (en) | Photovoltaic panel pollution degree state evaluation and cleaning decision system | |
| Peng et al. | Echo State Network ship motion modeling prediction based on Kalman filter | |
| CN114240687A (en) | Energy hosting efficiency analysis method suitable for comprehensive energy system | |
| Varkeshi et al. | Predicting the performance of Gorgan wastewater treatment plant using ANN-GA, CANFIS, and ANN models | |
| Dedeakayogullari et al. | The determination of mean and/or variance shifts with artificial neural networks | |
| CN110412878A (en) | The estimation of PID-RBF stable state and LSSVM hard measurement sewage BOD control method | |
| CN115618724A (en) | Thermal nonlinear system identification method, system, medium, equipment and terminal | |
| CN115458078A (en) | Multi-working-condition-based distributed output process compensation width migration modeling method | |
| CN116224774A (en) | Sludge drying treatment optimizing control method and system | |
| Teslyuk et al. | Synthesis of Neurocontroller for Intellectualization Tasks of Process Control Systems | |
| Jalalkamali et al. | Application of hybrid neural modeling and radial basis function neural network to estimate leakage rate in water distribution network | |
| CN110739030A (en) | Soft measurement method for small sample in ethylene production process | |
| Juuso et al. | Intelligent analysers and dynamic simulation in a biological water treatment process | |
| CN112816211A (en) | Belt conveyor fault diagnosis method based on PSO-BP algorithm | |
| Hu et al. | The soft measure model of dissolved oxygen based on RBF network in ponds | |
| Di et al. | Prediction of dissolved oxygen concentration based on improved RBF | |
| Hu et al. | Fault Diagnosis Using Deep Learning for Wastewater Treatment Processes |
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: 20201103 |
|
| RJ01 | Rejection of invention patent application after publication |