A MBBR pool sewage treatment method based on convolutional neural network

Abstract

The present invention provides a MBBR pool sewage treatment method based on convolutional neural network, which relates to the technical field of sewage treatment, including: collecting various water quality parameters and operating parameters in the sewage treatment process in the MBBR process; pre-processing the water quality parameters and operating parameters respectively to obtain pre-processing data; constructing a convolutional neural network model, and training the convolutional neural network model to obtain a trained sewage treatment model; inputting the sewage treatment parameters collected in real time into the sewage treatment model to obtain the sewage treatment results, and adjusting the operating parameters of the sewage treatment equipment in the MBBR process according to the sewage treatment results to achieve the best sewage treatment effect. The present invention uses convolutional neural networks to realize the intelligence of the sewage treatment process, which not only improves the accuracy of sewage treatment, but also greatly improves the efficiency of sewage treatment.

Description

Technical Field

[0001] The present invention relates to the technical field of sewage treatment, and in particular to a MBBR pool sewage treatment method based on convolutional neural network. Background Art

[0002] With the acceleration of industrialization and urbanization, the problem of sewage treatment is becoming increasingly serious.

[0003] MBBR technology is currently a popular technology in the field of sewage treatment. The principle of MBBR technology is to use the basic principle of biofilm method. By adding a certain amount of suspended carriers to the reactor, the biomass and biological species in the reactor are increased, thereby improving the treatment efficiency of the reactor. Since the density of the filler is close to that of water, it is completely mixed with water during aeration, and the environment for microbial growth is gas, liquid, and solid. The collision and shearing effect of the carrier in the water makes the air bubbles smaller and increases the utilization rate of oxygen. In addition, each carrier has different biological species inside and outside, with some anaerobic or facultative aerobic bacteria growing inside and aerobic bacteria outside. In this way, each carrier is a microreactor, allowing nitrification and denitrification reactions to exist simultaneously.

[0004] However, at present, the traditional sewage treatment method for MBBR process mainly relies on manual and experience-based control, which is time-consuming and labor-intensive, and it is impossible to monitor the sewage treatment effect in real time. Therefore, it is urgent to develop a real-time, efficient and accurate sewage treatment method. Summary of the invention

[0005] In order to overcome the shortcomings of the prior art, the object of the present invention is to provide a MBBR pool sewage treatment method based on convolutional neural network.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] A MBBR pool sewage treatment method based on convolutional neural network, comprising:

[0008] Collect various water quality parameters and operation parameters in the wastewater treatment process in the MBBR process; the water quality parameters include pH value, chemical oxygen demand, dissolved oxygen, biological oxygen demand, ammonia nitrogen and total phosphorus; the operation parameters include internal reflow set value, external reflow volume, residual sludge volume and aeration tank air volume;

[0009] Preprocessing the water quality parameters and the operating parameters respectively to obtain preprocessing data;

[0010] Constructing a convolutional neural network model, and training the convolutional neural network model through the pre-processed data to obtain a trained sewage treatment model; the input layer of the convolutional neural network model receives the pre-processed data, the hidden layer of the convolutional neural network model performs feature extraction and recognition, and the output layer of the convolutional neural network model outputs the sewage treatment result;

[0011] The sewage treatment parameters collected in real time are input into the sewage treatment model to obtain the sewage treatment results, and the operating parameters of the sewage treatment equipment in the MBBR process are adjusted according to the sewage treatment results to achieve the best sewage treatment effect.

[0012] Preferably, various water quality parameters and operating parameters in the wastewater treatment process in the MBBR process are collected, including:

[0013] During the sewage treatment process, a plurality of monitoring points are set; monitoring equipment is set at the monitoring points; the monitoring equipment is used to collect the water quality parameters;

[0014] A data collector is arranged on the sewage treatment equipment; the data collector is used to collect the operating parameters.

[0015] Preferably, the water quality parameter and the operating parameter are preprocessed respectively to obtain preprocessed data, including:

[0016] Performing denoising on the water quality parameters collected at each monitoring point to obtain denoised water quality parameters;

[0017] The denoised water quality parameters of each monitoring point within an adjacent preset range are merged to obtain the fused water quality parameters of the monitoring points;

[0018] Using a filter window to detect noise points on the operating parameters to obtain a noise mean;

[0019] When the noise mean in the filtering window is greater than a preset threshold, denoising the operating parameters in the corresponding filtering window;

[0020] Slide the filter window and return to the step of "when the noise mean in the filter window is greater than the preset threshold, denoising the operating parameters in the corresponding filter window" until the entire sequence of operating parameters is traversed to obtain the denoised operating parameters.

[0021] Preferably, the water quality parameters collected at each monitoring point are denoised to obtain the denoised water quality parameters, including:

[0022] Constructing a connection model according to the spatial distance between each of the monitoring points;

[0023] Constructing a data denoising function using the connection degree model;

[0024] The data denoising function is used to remove abnormal values ​​of the data information at each monitoring point to obtain denoised water quality parameters.

[0025] Preferably, the formula of the connection degree model is: Among them, σ represents the average value of the data value difference between the monitoring point at point m and the monitoring point at point n in the historical time period, dis(m,n) represents the spatial distance between the monitoring point at point m and the monitoring point at point n, and R represents the preset parameter.

[0026] Preferably, constructing a data denoising function using the connectivity model includes:

[0027] The denoising threshold is constructed using the weighted value of the data value of each monitoring point; the formula of the denoising threshold is: Among them, Y i represents the value of the ith data collected by the monitoring device, W i Represents Y i The weighted value of

[0028] A data denoising function is constructed based on the denoising threshold and the connection degree model; the calculation formula of the data denoising function is:

[0029]

[0030] in, represents the water quality parameter after denoising at the corresponding monitoring point at time t, x(m,t) represents the measurement value of the monitoring device at point m at time t, x(n,t) represents the measurement value of the monitoring device at point n at time t, ρ represents the adjustable threshold, | N m| Indicates the number of monitoring devices at the same monitoring point.

[0031] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0032] The present invention provides a MBBR pool sewage treatment method based on a convolutional neural network, including: collecting various water quality parameters and operating parameters in the sewage treatment process in the MBBR process; pre-processing the water quality parameters and the operating parameters respectively to obtain pre-processing data; constructing a convolutional neural network model, and training the convolutional neural network model through the pre-processing data to obtain a trained sewage treatment model; the input layer of the convolutional neural network model receives the pre-processing data, the hidden layer of the convolutional neural network model extracts and identifies features, and the output layer of the convolutional neural network model outputs the sewage treatment result; the sewage treatment parameters collected in real time are input into the sewage treatment model to obtain the sewage treatment result, and the operating parameters of the sewage treatment equipment in the MBBR process are adjusted according to the sewage treatment result to achieve the best sewage treatment effect. The present invention uses a convolutional neural network to realize the intelligence of the sewage treatment process, which not only improves the accuracy of sewage treatment, but also greatly improves the efficiency of sewage treatment. In addition, the present invention can also adjust the operating parameters of the sewage treatment equipment in real time, further improving the effect of sewage treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0034] Figure 1 A flow chart of a method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] The purpose of the present invention is to provide a MBBR pool sewage treatment method based on convolutional neural network, which realizes the intelligence of the sewage treatment process by using convolutional neural network, which not only improves the accuracy of sewage treatment, but also greatly improves the efficiency of sewage treatment, and can also adjust the operating parameters of sewage treatment equipment in real time, further improving the effect of sewage treatment.

[0037] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0038] Figure 1 A flow chart of a method provided by an embodiment of the present invention, such as Figure 1 As shown, the present invention provides a MBBR pool sewage treatment method based on a convolutional neural network, comprising:

[0039] Step 100: Collect various water quality parameters and operation parameters in the wastewater treatment process in the MBBR process; the water quality parameters include pH value, chemical oxygen demand, dissolved oxygen, biological oxygen demand, ammonia nitrogen and total phosphorus; the operation parameters include internal reflow set value, external reflow volume, residual sludge volume and aeration tank air volume.

[0040] Step 200: pre-processing the water quality parameter and the operation parameter respectively to obtain pre-processed data;

[0041] Step 300: construct a convolutional neural network model, and train the convolutional neural network model through the pre-processed data to obtain a trained sewage treatment model; the input layer of the convolutional neural network model receives the pre-processed data, the hidden layer of the convolutional neural network model performs feature extraction and recognition, and the output layer of the convolutional neural network model outputs the sewage treatment result;

[0042] Step 400: Input the real-time collected sewage treatment parameters into the sewage treatment model to obtain sewage treatment results, and adjust the operating parameters of the sewage treatment equipment in the MBBR process according to the sewage treatment results to achieve the best sewage treatment effect.

[0043] In this embodiment, the dissolved oxygen and internal reflow set values, external reflow volume, residual sludge volume, aeration tank blowing volume, etc. under the current process can be obtained through collection and calculation; then, the air volume distribution and dissolved oxygen control are executed by automatically adjusting the valve light operating equipment, so that the actual dissolved oxygen meets the set value requirements of the bioprocess intelligent optimization system and reduces fluctuations.

[0044] Wherein, step 100 includes:

[0045] During the sewage treatment process, a plurality of monitoring points are set; the monitoring equipment is used to collect the water quality parameters;

[0046] A data collector is arranged on the sewage treatment equipment; the data collector is used to collect the operating parameters.

[0047] Specifically, the monitoring equipment in this embodiment can be a water quality detector, which is arranged in a sewage treatment pipeline or a sewage treatment pool, and the water quality detector is arranged longitudinally along the pipeline or the treatment pool. The water quality detector needs to be arranged within the effective range, that is, each water quality detector can clearly identify the data value of the coverage area, and a new water quality detector needs to be added beyond the range. The water quality detector can be arranged on the left and right sides of the pipeline.

[0048] In addition, there may be multiple water quality detectors at each monitoring point. By integrating the acquisition parameters of water quality detectors in the same area, the confidence level of the acquired data can be further improved.

[0049] Furthermore, step 200 includes:

[0050] Performing denoising on the water quality parameters collected at each monitoring point to obtain denoised water quality parameters;

[0051] The denoised water quality parameters of each monitoring point within an adjacent preset range are merged to obtain the fused water quality parameters of the monitoring points;

[0052] Using a filter window to detect noise points on the operating parameters to obtain a noise mean;

[0053] When the noise mean in the filtering window is greater than a preset threshold, denoising the operating parameters in the corresponding filtering window;

[0054] Slide the filter window and return to the step of "when the noise mean in the filter window is greater than the preset threshold, denoising the operating parameters in the corresponding filter window" until the entire sequence of operating parameters is traversed to obtain the denoised operating parameters.

[0055] Optionally, this embodiment can use wavelet denoising to process the operating parameters. In the denoising process, the value of the wavelet threshold has an important influence on the final denoising effect. The wavelet threshold can separate the wavelet coefficients of the noise and the original signal to the greatest extent. As the number of wavelet decomposition layers increases, the wavelet coefficient area where the noise is located gradually decreases, and most of the decomposed wavelet coefficients come from the original signal. Therefore, this embodiment can construct a self-transformation filtering threshold based on the decomposition scale.

[0056] Furthermore, the filtering process of constructing the self-transformation filtering threshold based on the decomposition scale is as follows:

[0057] Performing wavelet decomposition on the operating parameters at multiple scales to obtain corresponding wavelet coefficients;

[0058] constructing a filtering threshold according to the sequence size and decomposition scale of the operating parameters;

[0059] constructing a filtering model using the filtering threshold;

[0060] The operating parameters are filtered using the filtering model to obtain filtered operating parameters.

[0061] Preferably, the water quality parameters collected at each monitoring point are denoised to obtain the denoised water quality parameters, including:

[0062] Constructing a connection model according to the spatial distance between each of the monitoring points;

[0063] Constructing a data denoising function using the connection degree model;

[0064] The data denoising function is used to remove abnormal values ​​of the data information at each monitoring point to obtain denoised water quality parameters.

[0065] Preferably, the formula of the connection degree model is: Among them, σ represents the average value of the data value difference between the monitoring point at point m and the monitoring point at point n in the historical time period, dis(m,n) represents the spatial distance between the monitoring point at point m and the monitoring point at point n, and R represents the preset parameter.

[0066] Specifically, in this embodiment, the monitoring devices at the monitoring points at point m and point n are neighboring monitoring devices. The closer the positions of the two monitoring devices are, the greater the correlation between the two monitoring devices. Therefore, the present invention measures the degree of connection between the two monitoring devices based on the spatial distance between them. In practical applications, since the monitoring device may be affected by its own parameters or environmental factors, the measured value collected by the monitoring device at a certain moment may deviate greatly from the actual value. Therefore, the present invention introduces R to correct the problem of deviation in the correlation of the monitoring device due to the influence of its own parameters or environmental factors. In general, the value of R is 1.

[0067] Preferably, constructing a data denoising function using the connectivity model includes:

[0068] The denoising threshold is constructed using the weighted value of the data value of each monitoring point; the formula of the denoising threshold is: Among them, Y i represents the value of the ith data collected by the monitoring device, W i Represents Y i The weighted value of

[0069] A data denoising function is constructed based on the denoising threshold and the connection degree model; the calculation formula of the data denoising function is:

[0070]

[0071] in, represents the water quality parameter after denoising at the corresponding monitoring point at time t, x(m,t) represents the measurement value of the monitoring device at point m at time t, x(n,t) represents the measurement value of the monitoring device at point n at time t, ρ represents the adjustable threshold, | N m| Indicates the number of monitoring devices at the same monitoring point.

[0072] Furthermore, this embodiment can avoid a situation where a large deviation occurs between the measurement value collected by the monitoring device at a certain moment and the actual value by fusing the measurement values ​​of the monitoring device at the same monitoring point using a data denoising function, thereby making the measurement value of the sensor more realistic.

[0073] In addition, this embodiment uses monitoring equipment to collect data, and uses a fusion gateway to transmit the collected data. The fusion gateway can solve the data aggregation, access and processing problems of various types of sensing detection units, and realize unified aggregation processing and standardization. The fusion gateway supports a variety of uplink and downlink communication interfaces and powerful AT edge computing capabilities, and supports software and hardware encryption functions to ensure data security and transmission security. The system provides solutions based on OPGW optical fiber transmission, Beidou-3 short message communication, 4G / 5G transmission, 5.8G wireless self-organizing network transmission, etc., which can be flexibly configured according to the on-site environment, and supports multi-channel sub-group detection and switching functions. Various sensor data are integrated by the sewage treatment area to be tested to achieve accurate processing and transmission of multiple data.

[0074] Furthermore, the convolutional neural network of this embodiment also includes a fully connected layer. The hidden layer is composed of one or more convolutional layers and one or more pooling layers arranged at intervals; this embodiment implements a complete convolution implementation system from the input layer to the convolutional layer to the pooling layer to the fully connected layer to the output layer.

[0075] The beneficial effects of the present invention are as follows:

[0076] The present invention utilizes the powerful learning and recognition capabilities of convolutional neural networks to realize the intelligentization of the sewage treatment process, which not only improves the accuracy of sewage treatment, but also greatly improves the efficiency of sewage treatment. In addition, the present invention can also adjust the operating parameters of sewage treatment equipment in real time, further improving the effect of sewage treatment.

[0077] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0078] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A MBBR pool sewage treatment method based on convolutional neural network, characterized in that: include: Collect various water quality parameters and operation parameters in the wastewater treatment process in the MBBR process; the water quality parameters include pH value, chemical oxygen demand, dissolved oxygen, biological oxygen demand, ammonia nitrogen and total phosphorus; the operation parameters include internal reflow set value, external reflow volume, residual sludge volume and aeration tank air volume; The collected data is transmitted by using a fusion gateway, which is used for flexible configuration according to the on-site environment and supports multi-channel sub-group detection and switching functions. Various sensor data are integrated from the sewage treatment area to be tested to achieve accurate processing and transmission of multiple data; Preprocessing the water quality parameters and the operating parameters respectively to obtain preprocessing data; Constructing a convolutional neural network model, and training the convolutional neural network model through the pre-processed data to obtain a trained sewage treatment model; the input layer of the convolutional neural network model receives the pre-processed data, the hidden layer of the convolutional neural network model performs feature extraction and recognition, and the output layer of the convolutional neural network model outputs the sewage treatment result; Inputting the real-time collected sewage treatment parameters into the sewage treatment model to obtain sewage treatment results, and adjusting the operating parameters of the sewage treatment equipment in the MBBR process according to the sewage treatment results to achieve the best sewage treatment effect; Collect various water quality parameters and operation parameters during the wastewater treatment process in the MBBR process, including: During the sewage treatment process, a plurality of monitoring points are set; monitoring equipment is set at the monitoring points; the monitoring equipment is used to collect the water quality parameters; A data collector is arranged on the sewage treatment equipment; the data collector is used to collect the operating parameters; Preprocessing the water quality parameters and the operating parameters respectively to obtain preprocessing data includes: Performing denoising on the water quality parameters collected at each monitoring point to obtain denoised water quality parameters; The denoised water quality parameters of each monitoring point within an adjacent preset range are merged to obtain the fused water quality parameters of the monitoring points; Using a filter window to detect noise points on the operating parameters to obtain a noise mean; When the noise mean in the filtering window is greater than a preset threshold, denoising the operating parameters in the corresponding filtering window; Slide the filter window and return to step "when the noise mean in the filter window is greater than the preset threshold, denoise the operating parameters in the corresponding filter window" until the entire sequence of operating parameters is traversed to obtain the denoised operating parameters; The process of filtering based on the self-transformation filtering threshold constructed based on the decomposition scale is as follows: Performing wavelet decomposition on the operating parameters at multiple scales to obtain corresponding wavelet coefficients; constructing a filtering threshold according to the sequence size and decomposition scale of the operating parameters; constructing a filtering model using the filtering threshold; Filtering the operating parameters using the filtering model to obtain filtered operating parameters; The water quality parameters collected at each monitoring point are subjected to denoising to obtain denoised water quality parameters, including: Constructing a connection model according to the spatial distance between each of the monitoring points; Constructing a data denoising function using the connection degree model; The data denoising function is used to remove abnormal values ​​of the data information at each monitoring point to obtain denoised water quality parameters; The formula of the connection degree model is: ;in, It represents the average value of the data value difference between the monitoring point at point m and the monitoring point at point n in the historical time period. represents the spatial distance between the monitoring point at point m and the monitoring point at point n, Indicates preset parameters; The data denoising function is constructed using the connection degree model, including: The denoising threshold is constructed using the weighted value of the data value of each monitoring point; the formula of the denoising threshold is: ;in, represents the value of the i-th data collected by the monitoring device, express The weighted value of; constructing a data denoising function based on the denoising threshold and the connection degree model; the calculation formula of the data denoising function is: ;in, represents the water quality parameters after denoising at the corresponding monitoring point at time t, represents the measured value of the monitoring device at point m at time t, represents the measurement value of the monitoring device at point n at time t, represents an adjustable threshold, Indicates the number of monitoring devices at the same monitoring point. .

Images