Power transmission and transformation project wastewater purification treatment method and system based on neural network

Abstract

The invention discloses a power transmission and transformation project wastewater purification treatment method and system based on a neural network. The method comprises the following steps: detecting first particle size distribution information and a first near-infrared absorption spectrum of pollutants in wastewater to be treated; according to the first near-infrared absorption spectrum, a first corresponding relation between the particle size of the oil droplet and a wavelength absorption peak, and a second corresponding relation between the particle size of the suspended matter and the wavelength absorption peak, identifying the particle sizes of the oil droplet and the suspended matter according to the first particle size distribution information; determining purification parameters through a first deep neural network model according to the identified first oil droplet particle size distribution information and the first suspended matter particle size distribution information; controlling a flocculating agent adding pump to add a flocculating agent according to the target flocculating agent type and the first adding amount of the purification parameters, and configuring a pressure parameter of a bubble generator to be a pressure parameter corresponding to a first bubble particle size so as to generate bubbles of the first bubble particle size, so that wastewater purification treatment is realized; according to the invention, the purification effect of the wastewater containing oil and suspended solids in the power transmission and transformation project can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of purification treatment, and in particular to a method and system for purifying wastewater in a power transmission and transformation project based on a neural network. Background Art

[0002] During the implementation of the power transmission and transformation project, the wastewater generated during the flushing process of the mechanical equipment used in the base excavation is rich in high-concentration suspended solids and high-concentration grease, and has the characteristics of complex water quality, high chemical oxygen demand and biochemical oxygen demand, poor oil-water separation, large treatment scale, and difficult removal. If the wastewater is discharged directly, it will affect the regional water quality and have an adverse impact on the ecological environment.

[0003] Flocculants play a key role in the treatment of wastewater containing oil and suspended matter generated during the implementation of power transmission and transformation projects. However, there are many types of flocculants, ranging from low molecular weight to high molecular weight, from single type to composite type; flocculants can be generally divided into two categories according to their chemical composition: inorganic flocculants and organic flocculants. Among them, inorganic flocculants include inorganic coagulants and inorganic polymer flocculants; organic flocculants include synthetic organic polymer flocculants, natural organic polymer flocculants and microbial flocculants. Different flocculants have different decontamination effects. The molecular weight and charge density of the flocculant will significantly affect the surface charge density of the flocculant, and then affect the adhesion of bubbles on the surface of the flocculant. The strength of the adhesion of the flocculant surface directly determines the decontamination effect of the flocculant. At present, for the treatment of wastewater containing oil and suspended matter generated during the implementation of power transmission and transformation projects, a single type of flocculant is mainly used to be added to the wastewater in a fixed proportion for decontamination. The characteristics of the wastewater containing oil and suspended matter generated during the implementation of power transmission and transformation projects are not considered, and the purification effect is poor. Summary of the invention

[0004] In view of the problems existing in the prior art, the embodiments of the present invention provide a method and system for purifying wastewater from power transmission and transformation projects based on a neural network, which can effectively improve the purification effect of wastewater containing oil and suspended matter generated during the flushing process of mechanical equipment in power transmission and transformation projects.

[0005] In a first aspect, an embodiment of the present invention provides a method for purifying wastewater in a power transmission and transformation project based on a neural network, comprising: Detecting the first particle size distribution information of pollutants in the wastewater to be treated in the power transmission and transformation project by using a laser particle size detector; Acquiring a first near-infrared absorption spectrum of the wastewater to be treated; According to the first near-infrared absorption spectrum and the pre-established first correspondence between the oil droplet particle size and the wavelength absorption peak, and the pre-established second correspondence between the suspended matter particle size and the wavelength absorption peak, the pollutants corresponding to different particle sizes in the first particle size distribution information are identified as oil droplets and suspended matter, to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; According to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, purification parameters are determined by a pre-established first deep neural network model; wherein the purification parameters include a target flocculant type, a first dosage, and a first bubble particle size; The flocculant dosing pump is controlled to add the corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage, and the pressure parameter of the bubble generator is configured to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated to purify the wastewater to be treated.

[0006] As an improvement of the above solution, the method further includes: Real-time acquisition of a second near-infrared absorption spectrum during the purification process of the wastewater to be treated; Real-time acquisition of second particle size distribution information during the purification process of the wastewater to be treated; According to the second near-infrared absorption spectrum and the first corresponding relationship between the oil droplet diameter and the wavelength absorption peak, the particle diameter matching the oil droplet diameter and the concentration information of the corresponding particle diameter are screened out from the second particle diameter distribution information to obtain the second oil droplet diameter distribution information in the wastewater purification process to be treated; The bubbles in the wastewater to be treated are monitored in real time by a high-speed camera, and the bubble images monitored in real time are recognized to obtain the bubble particle size distribution information in the wastewater to be treated; According to the second oil droplet particle size distribution information and the bubble particle size distribution information, a purification adjustment parameter is obtained through a pre-established second deep neural network model; wherein the purification adjustment parameter includes a second dosage of a flocculant corresponding to a target flocculant type and a second bubble particle size; The flocculant dosing pump is adjusted to add a corresponding flocculant dosage to the wastewater to be treated according to the second dosage, and the bubble generator is controlled to input bubbles of the second bubble size into the wastewater to be treated, so as to purify and optimize the wastewater to be treated.

[0007] As an improvement of the above solution, the method further includes: Obtaining third near-infrared absorption spectra of multiple oily wastewater samples containing different oil droplet sizes; Determine the wavelength absorption peaks corresponding to the different oil droplet sizes according to the third near-infrared absorption spectra of the different oil droplet sizes, and establish a first corresponding relationship between the oil droplet size and the wavelength absorption peak; obtaining fourth near infrared absorption spectra of a plurality of suspended matter wastewater samples having different suspended matter particle sizes; According to the fourth near-infrared absorption spectra of different suspended matter particle sizes, wavelength absorption peaks corresponding to different suspended matter particle sizes are determined, and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak is established.

[0008] As an improvement of the above scheme, according to the first near-infrared absorption spectrum and the pre-established first correspondence between the oil droplet particle size and the wavelength absorption peak, and the second correspondence between the suspended matter particle size and the wavelength absorption peak, the pollutants corresponding to different particle sizes in the first particle size distribution information are identified with oil droplets and suspended matter to obtain the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, including: According to the first near-infrared absorption spectrum and the first corresponding relationship between the oil droplet diameter and the wavelength absorption peak, the target oil droplet diameter in the wastewater to be treated is obtained; Screening out the particle size matching the target oil droplet particle size and the concentration information of the corresponding particle size from the first particle size distribution information to obtain the first oil droplet particle size distribution information; According to the first near-infrared absorption spectrum and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak, the target suspended matter particle size in the wastewater to be treated is obtained; The particle size matching the target suspended matter particle size and the concentration information of the corresponding particle size are screened out from the first particle size distribution information to obtain the first suspended matter particle size distribution information.

[0009] As an improvement of the above solution, the purification parameters are determined according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information through a pre-established first deep neural network model, including: Screening out at least one candidate flocculant type from a flocculant database according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information; Constructing a wastewater feature vector according to the first oil droplet particle size distribution information, the first suspended matter particle size distribution information, and the current pH value and current water temperature of the wastewater to be treated; According to the molecular weight and charge density of the flocculant corresponding to at least one candidate flocculant type, construct a flocculant feature vector of the corresponding candidate flocculant type; Inputting the wastewater feature vector and the flocculant feature vectors corresponding to each of the candidate flocculant types into the first deep neural network model to obtain a flocculant particle size change graph of the flocculant corresponding to each of the candidate flocculant types in the wastewater to be treated; The purification parameters are determined according to the flocculent particle size variation diagram corresponding to each of the candidate flocculant types.

[0010] As an improvement of the above solution, the method of determining the purification parameters according to the flocculant particle size change diagram corresponding to each candidate flocculant type includes: Calculating the flocculation efficiency and average flocculent particle size of each candidate flocculant type according to the flocculent particle size change diagram corresponding to each candidate flocculant type; Select the candidate flocculant type corresponding to the highest flocculation efficiency as the target flocculant type; Calculating a first dosage of a flocculant corresponding to the target flocculant type according to the flocculation efficiency of the target flocculant type; The first bubble particle size is calculated according to the charge density of the flocculant corresponding to the target flocculant type and the average flocculent particle size.

[0011] As an improvement of the above solution, the first bubble particle size is calculated according to the charge density of the flocculant corresponding to the target flocculant type, including: Calculating a target bubble attachment degree under the target flocculant type according to the charge density and average flocculent particle size of the flocculant corresponding to the target flocculant type; Obtaining the first bubble particle size according to the target bubble attachment degree and a pre-established first mapping table between the bubble attachment degree and the bubble particle size; The pressure parameter of the bubble generator is configured as a pressure parameter corresponding to the first bubble particle size, so that the bubble generator generates bubbles of the first bubble particle size.

[0012] As an improvement of the above solution, the purification adjustment parameters are obtained by using a pre-established second deep neural network model according to the second oil droplet size distribution information and the bubble size distribution information, including: Calculating the oil droplet size distribution width at the current moment according to the second oil droplet size distribution information at the current moment; When the oil droplet particle size distribution width is greater than the preset alarm particle size range, the purification adjustment parameters are obtained according to the second oil droplet particle size distribution information at the current moment and the bubble particle size distribution information at the current moment through a pre-established second deep neural network model.

[0013] As an improvement of the above scheme, the purification adjustment parameters are obtained by using a pre-established second deep neural network model according to the second oil droplet particle size distribution information at the current moment and the bubble particle size distribution information at the current moment, including: Constructing an oil droplet size distribution feature vector according to the second oil droplet size distribution information at the current moment; According to the bubble particle size distribution information at the current moment, a bubble particle size distribution feature vector is constructed; Inputting the oil droplet size distribution feature vector and the bubble size distribution feature vector into the second deep neural network model to obtain a predicted bubble adhesion degree; According to the predicted bubble adhesion degree and the target bubble adhesion degree, a second bubble particle size and a second dosage are calculated.

[0014] In a second aspect, an embodiment of the present invention provides a power transmission and transformation engineering wastewater purification system based on a neural network, comprising: A first particle size detection module is used to detect first particle size distribution information of pollutants in wastewater to be treated in a power transmission and transformation project through a laser particle size detector; A first near-infrared absorption spectrum acquisition module, used to acquire a first near-infrared absorption spectrum of the wastewater to be treated; A first particle size detection module is used to detect first particle size distribution information of pollutants in wastewater to be treated in a power transmission and transformation project through a laser particle size detector; A first near-infrared absorption spectrum acquisition module, used to acquire a first near-infrared absorption spectrum of the wastewater to be treated; a first particle size identification module, for identifying oil droplets and suspended matter for pollutants corresponding to different particle sizes in the first particle size distribution information according to the first near-infrared absorption spectrum and a pre-established first correspondence between oil droplet particle size and wavelength absorption peak, and a pre-established second correspondence between suspended matter particle size and wavelength absorption peak, to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; A purification parameter determination module, used to determine purification parameters according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information through a pre-established first deep neural network model; wherein the purification parameters include a target flocculant type, a first dosage, and a first bubble particle size; The first control module is used to control the flocculant dosing pump to add the corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage, and configure the pressure parameter of the bubble generator to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated to purify the wastewater to be treated.

[0015] Compared with the prior art, a method and system for purifying wastewater from a power transmission and transformation project based on a neural network in an embodiment of the present invention detects first particle size distribution information of pollutants in wastewater to be treated in the power transmission and transformation project by a laser particle size detector, and simultaneously obtains a first near-infrared absorption spectrum of the wastewater to be treated; then, based on the first near-infrared absorption spectrum and a pre-established first correspondence between oil droplet particle size and wavelength absorption peak, and a second correspondence between suspended matter particle size and wavelength absorption peak, oil droplets and suspended matter are identified for pollutants corresponding to different particle sizes in the first particle size distribution information, to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; based on the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, the oil droplets and suspended matter are identified for pollutants corresponding to different particle sizes in the first particle size distribution information. distribution information, and determine the purification parameters through a pre-established first deep neural network model; finally, control the flocculant dosing pump to add the corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage indicated by the purification parameters, and at the same time configure the pressure parameter of the bubble generator to the pressure parameter corresponding to the first bubble particle size indicated by the purification parameter, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated, and through the joint action of the flocculant and the bubbles, the grease and suspended matter in the wastewater are flocculated and floated, so as to achieve the separation of grease and suspended matter in the wastewater, realize the purification of the wastewater to be treated, and improve the purification effect of the oil-containing and suspended matter wastewater generated during the flushing process of mechanical equipment in the power transmission and transformation project. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solution of the present invention, the drawings used in the implementation mode will be briefly introduced below. Obviously, the drawings described below are only some implementation modes of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0017] Figure 1 It is a flow chart of a method for purifying wastewater in a power transmission and transformation project based on a neural network provided by an embodiment of the present invention; Figure 2 is a schematic block diagram of a wastewater treatment process provided by an embodiment of the present invention; Figure 3 It is a schematic diagram of a process for determining purification parameters provided by an embodiment of the present invention; Figure 4 It is a structural block diagram of a wastewater purification system for power transmission and transformation projects based on a neural network provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] 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.

[0019] It is understood that the various numbers involved in the embodiments of the present invention are only for the convenience of description and are not intended to limit the scope of the present application. The size of the sequence number of each process does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic.

[0020] In embodiments of the present invention, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. The terms "include", "comprises", or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article, or device. In the absence of further restrictions, the elements defined by the statement "includes..." do not exclude the presence of other identical elements in the process, method, article, or device including the elements. The term "plurality or several" refers to two or more.

[0021] See also Figure 1 , Figure 1 The flowchart of a method for purifying wastewater from power transmission and transformation projects based on a neural network provided by an embodiment of the present invention is as follows. The method for purifying wastewater from power transmission and transformation projects based on a neural network specifically includes: S11: Detecting first particle size distribution information of pollutants in wastewater to be treated in a power transmission and transformation project by using a laser particle size detector; S12: Obtaining a first near-infrared absorption spectrum of the wastewater to be treated; S13: performing oil droplet and suspended matter identification on pollutants corresponding to different particle sizes in the first particle size distribution information according to the first near-infrared absorption spectrum and the pre-established first correspondence between the oil droplet particle size and the wavelength absorption peak, and the second correspondence between the suspended matter particle size and the wavelength absorption peak, to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; The specific process of oil droplet and suspended matter identification is as follows: According to the first near-infrared absorption spectrum and the first corresponding relationship between the oil droplet diameter and the wavelength absorption peak, the target oil droplet diameter in the wastewater to be treated is obtained; Screening out the particle size matching the target oil droplet particle size and the concentration information of the corresponding particle size from the first particle size distribution information to obtain the first oil droplet particle size distribution information; According to the first near-infrared absorption spectrum and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak, the target suspended matter particle size in the wastewater to be treated is obtained; The particle size matching the target suspended matter particle size and the concentration information of the corresponding particle size are screened out from the first particle size distribution information to obtain the first suspended matter particle size distribution information.

[0022] S14: Determine purification parameters through a pre-established first deep neural network model according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information; wherein the purification parameters include a target flocculant type, a first dosage, and a first bubble particle size; S15: Control the flocculant dosing pump to add corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage, and configure the pressure parameter of the bubble generator to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated, so as to purify the wastewater to be treated.

[0023] It should be noted that the neural network-based power transmission and transformation project wastewater purification treatment method described in the embodiment of the present invention can be executed by a server, computer and other terminal devices equipped with a neural network-based power transmission and transformation project wastewater purification treatment system. The system is respectively communicated with a laser particle size detector for real-time monitoring of the particle size distribution of pollutants (such as oil, suspended matter) in the wastewater to be treated, and a near-infrared spectrometer for real-time monitoring of the near-infrared absorption spectrum of the wastewater to be treated. At the same time, the system is also communicated with a flocculant dosing pump and a bubble generator to control the flocculant and the dosage added by the flocculant dosing pump to the wastewater to be treated, and to control the pressure parameters of the bubble generator to generate bubbles of suitable bubble size to the wastewater to be treated. The combined action of the flocculant and the bubbles makes the wastewater flocculate and float, and the suspended matter and oil in the wastewater are separated, so as to purify the wastewater to be treated.

[0024] For example, Figure 2As shown, the wastewater to be treated in the power transmission and transformation project can be pressurized by the delivery pump device 1 to be delivered to the wastewater purification area 2 for storage, and then the laser particle size detector 3 is started to analyze the pollutant particle size of the wastewater to be treated in the wastewater purification area to obtain the first particle size distribution information and upload it to the system synchronously; at the same time, the near-infrared spectrometer 4 is started to detect the near-infrared absorption spectrum of the wastewater to be treated to obtain the first near-infrared absorption spectrum and upload it to the system synchronously; after the system receives the first particle size distribution information and the first near-infrared absorption spectrum, it first calculates the particle size distribution information based on the first corresponding relationship between the oil droplet particle size and the wavelength absorption peak stored locally, and the suspended solids. The second correspondence between the particle size and the wavelength absorption peak is used to perform wavelength absorption peak matching analysis on the first near-infrared absorption spectrum, and the particle size and concentration information matching the oil droplet particle size in the first particle size distribution information are identified to obtain the first oil droplet particle size distribution information. At the same time, the particle size and concentration information matching the suspended matter particle size in the first particle size distribution information are identified to obtain the first suspended matter particle size distribution information. Afterwards, according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, the first deep neural network model is used to determine the purification including the target flocculant type, the first dosage, and the first bubble particle size. Parameters; based on the obtained purification parameters, the system controls the flocculant dosing pump 5 arranged in the wastewater purification area to select the flocculant corresponding to the target flocculant type and add it to the wastewater to be treated according to the first dosage, and at the same time configures the pressure parameter of the bubble generator 6 arranged in the wastewater purification area to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator generates bubbles of the first bubble particle size to the wastewater to be treated, and the suspended matter and grease in the wastewater are flocculated and floated by the combined action of the flocculant and the bubbles, so as to achieve the separation of grease and suspended matter in the wastewater, and realize the purification of the wastewater to be treated. The embodiment of the present invention fully considers the particle size distribution of grease and suspended matter in the oil-containing and suspended matter wastewater generated during the flushing process of mechanical equipment in power transmission and transformation projects, and combines deep neural networks to control the type, dosage, and bubble particle size of flocculants, so as to obtain purification parameters that are suitable for the particle size distribution of grease and suspended matter in the oil-containing and suspended matter wastewater of the current power transmission and transformation projects. Compared with the prior art that uses a single type of flocculant to be added to the wastewater in a fixed proportion for decontamination, the embodiment of the present invention can effectively improve the purification effect of the oil-containing and suspended matter wastewater generated during the flushing process of mechanical equipment in power transmission and transformation projects.

[0025] Furthermore, the method further comprises: Obtaining third near-infrared absorption spectra of multiple oily wastewater samples containing different oil droplet sizes; Determine the wavelength absorption peaks corresponding to the different oil droplet sizes according to the third near-infrared absorption spectra of the different oil droplet sizes, and establish a first corresponding relationship between the oil droplet size and the wavelength absorption peak; obtaining fourth near infrared absorption spectra of a plurality of suspended matter wastewater samples having different suspended matter particle sizes; According to the fourth near-infrared absorption spectra of different suspended matter particle sizes, wavelength absorption peaks corresponding to different suspended matter particle sizes are determined, and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak is established.

[0026] The embodiment of the present invention uses a near-infrared spectrometer to scan the oily wastewater sample and the suspended matter wastewater sample respectively using a near-infrared light wave scan of 900 to 1700nm. Then, based on the third near-infrared absorption spectrum of the oily wastewater sample, the system analyzes the wavelength absorption peaks of different oil droplet sizes. The absorption intensities of different wavelengths of light for oil droplets of different particle sizes are different, and the peak position of the spectrum is positively correlated with the oil droplet size. For example, at a wavelength of 1200nm, oil droplets with a particle size of 20 microns have the strongest absorption peak, and the absorbance reaches 0.85. As the oil droplet size decreases, the absorption peak shifts toward the short wavelength direction. Based on different oil droplet sizes and corresponding wavelength absorption peaks, a first corresponding relationship is established; based on the fourth near-infrared absorption spectrum of the suspended matter wastewater sample, the wavelength absorption peaks of different suspended matter particle sizes are analyzed, and similarly, a second corresponding relationship between different suspended matter particle sizes and corresponding wavelength absorption peaks is established; and then the first corresponding relationship and the second corresponding relationship are stored in the local memory of the server.

[0027] It should be noted that the embodiments of the present invention do not specifically limit the manner of establishing the first correspondence and the second correspondence. For example, for different detected oil droplet diameters / suspended matter diameters, the maximum and minimum values ​​of the absorption peaks of the central wavelengths of the corresponding oil droplet diameters / suspended matter diameters under multiple samples can be selected to obtain the value range of the absorption peaks of the corresponding oil droplet diameters / suspended matter diameters, and the first correspondence / second correspondence can be established based on the different oil droplet diameters / suspended matter diameters and the corresponding absorption peak value ranges; or, the different detected oil droplet diameters / suspended matter diameters are sorted in order from small to large, and then evenly divided into a number of oil droplet size intervals / suspended particle size intervals, and the maximum and minimum values ​​of the absorption peaks of the central wavelengths of the corresponding oil droplet diameter intervals / suspended particle size intervals under multiple samples are selected to obtain the value range of the absorption peaks of the corresponding oil droplet diameter intervals / suspended particle size intervals, and the first correspondence / second correspondence can be established based on the different oil droplet diameters / suspended matter diameters and the corresponding absorption peak value ranges.

[0028] like Figure 3 As shown, S14: according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, a purification parameter is determined by a pre-established first deep neural network model, including: S141: Screening out at least one candidate flocculant type from a flocculant database according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information; The embodiment of the present invention establishes a flocculant database for storing different types of flocculants and the molecular weight and charge density of each flocculant, as well as the applicable oil droplet particle size range and the applicable suspended matter particle size range of each flocculant.

[0029] Before purifying the wastewater, the oil droplet particle size distribution width in the wastewater to be treated is calculated based on the oil droplet particle size in the first oil droplet particle size distribution information. For example, the maximum oil droplet particle size in the first oil droplet particle size distribution information is selected as the upper limit value and the minimum oil droplet particle size is selected as the lower limit value to obtain the oil droplet particle size distribution width. Similarly, the suspended matter particle size distribution width in the wastewater to be treated can be calculated.

[0030] Then, the first matching degree between the oil droplet size distribution width and the applicable oil droplet size range of each flocculant in the flocculant database is calculated. For example, the similarity between the oil droplet size distribution width and the applicable oil droplet size range of each flocculant is calculated by Euclidean distance, which is used as the first matching degree between the oil droplet size distribution width and the applicable oil droplet size range of each flocculant. Similarly, the second matching degree between the suspended matter particle size distribution width and the applicable suspended matter particle size range of each flocculant can be calculated. Then, the flocculant types whose first matching degree and second matching degree both exceed the preset matching degree threshold are selected as candidate flocculant types to achieve initial screening of flocculants.

[0031] S142: constructing a wastewater feature vector according to the first oil droplet particle size distribution information, the first suspended matter particle size distribution information, and the current pH value and current water temperature of the wastewater to be treated; In the embodiment of the present invention, the current pH value of the wastewater to be treated is detected by a pH meter, and the current water temperature of the wastewater to be treated is detected by a temperature sensor. The system constructs a one-dimensional wastewater feature vector based on the oil droplet size and corresponding concentration information in the first oil droplet size distribution information, the suspended matter size and corresponding concentration information in the first suspended matter size distribution information, the current pH value, and the current water temperature. For example, the wastewater feature vector Q= , represents the i-th oil droplet size in the first oil droplet size distribution information, represents the concentration of the i-th oil droplet size in the first oil droplet size distribution information, represents the i-th suspended matter particle size in the first suspended matter particle size distribution information, represents the concentration of the i-th suspended matter particle size in the first suspended matter particle size distribution information, Indicates the current pH value of the wastewater to be treated, Indicates the current water temperature of the wastewater to be treated.

[0032] S143: constructing a flocculant feature vector of the corresponding candidate flocculant type according to the molecular weight and charge density of the flocculant corresponding to at least one candidate flocculant type; In the embodiment of the present invention, a one-dimensional flocculant feature vector is constructed based on the molecular weight and charge density of the flocculant corresponding to each candidate flocculant type, for example, the flocculant feature vector of the jth candidate flocculant is: , represents the molecular weight of the jth candidate flocculant, represents the charge density of the jth candidate flocculant.

[0033] S144: Inputting the wastewater feature vector and the flocculant feature vectors corresponding to each of the candidate flocculant types into the first deep neural network model to obtain a flocculant particle size change graph of the flocculants corresponding to each of the candidate flocculant types in the wastewater to be treated.

[0034] It should be understood that the particle size distribution of flocs changes all the time during the flocculation process. Generally speaking, after entering the flocculation stage, the number of small-sized particles decreases rapidly, while the number of large particles increases significantly. The particle size of flocs gradually develops from a high peak to a low peak. At the same time, due to the agglomeration between particles, the particle size range of flocs gradually increases, that is, the peak width gradually increases. Based on the above principles, the embodiment of the present invention collects multiple oil-containing and suspended matter wastewater samples, and records the characteristic parameters of different oil-containing and suspended matter wastewater samples, including oil droplet size distribution, suspended matter size distribution, pH value, water temperature, and the changes in particle size distribution of flocs in wastewater samples over time after adding different types of flocculants. Then, according to the preset time interval (for example, 50s), the particle size distribution at multiple moments after the flocculant is added into the oil-containing and suspended matter wastewater sample is selected; for the particle size distribution at each moment, the flocculent particle size in the particle size distribution is used as the horizontal coordinate, and the corresponding number of particles is used as the vertical coordinate, and curve fitting is performed to obtain the flocculent particle size change curve of the corresponding virtual agent; then, the flocculent particle size change curves of the corresponding virtual agent at multiple moments are integrated to generate the flocculent particle size change graph of the corresponding virtual agent.

[0035] The characteristic parameters of the oily and suspended matter wastewater sample and the molecular weight and charge density of the flocculant put into the oily and suspended matter wastewater sample are used as model inputs, and the flocculant particle size change diagram of the oily and suspended matter wastewater sample is used as label data. The first deep neural network model is trained until the preset model accuracy requirement is met, and the trained first deep neural network model is obtained and stored locally in the system for subsequent call at any time. It should be noted that the model training process of the first deep neural network model belongs to the prior art and is not described in detail here. Through the model training of the first deep neural network model, the flocculant particle size changes of the corresponding flocculants under different wastewater environments and flocculant molecular weights and electrical load densities can be mined, so as to facilitate the screening of candidate target flocculant types that are suitable for the current wastewater to be treated.

[0036] The embodiment of the present invention predicts the change of flocculant particle size by inputting the wastewater feature vector and the flocculant feature vector corresponding to any candidate flocculant type into the above-mentioned trained first deep neural network model, and outputs a flocculant particle size change graph of the candidate flocculant type.

[0037] Among them, the floc particle size change diagram records the floc particle size change curve at different times (for example, after entering the flocculation stage, 50s, 100s, 150s, 200s, 250s, etc.), and the floc particle size change curve describes the change trend of the floc particle size and the number of particles.

[0038] It should be noted that the first deep neural network model can be constructed using neural networks such as convolutional neural network CNN and recurrent neural network RNN, which is not specifically limited in the embodiments of the present invention.

[0039] S145: Determine the purification parameters according to the flocculant particle size variation diagram corresponding to each of the candidate flocculant types.

[0040] Specifically, the step of determining the purification parameters according to the flocculant particle size variation diagram corresponding to each candidate flocculant type includes: Calculating the flocculation efficiency and average flocculent particle size of each candidate flocculant type according to the flocculent particle size change diagram corresponding to each candidate flocculant type; Select the candidate flocculant type corresponding to the highest flocculation efficiency as the target flocculant type; Calculating a first dosage of a flocculant corresponding to the target flocculant type according to the flocculation efficiency of the target flocculant type; The first bubble particle size is calculated according to the charge density of the flocculant corresponding to the target flocculant type and the average flocculent particle size.

[0041] Specifically, the first bubble particle size is calculated according to the charge density of the flocculant corresponding to the target flocculant type, including: Calculating a target bubble attachment degree under the target flocculant type according to the charge density and average flocculent particle size of the flocculant corresponding to the target flocculant type; Obtaining the first bubble particle size according to the target bubble attachment degree and a pre-established first mapping table between the bubble attachment degree and the bubble particle size; The pressure parameter of the bubble generator is configured as a pressure parameter corresponding to the first bubble particle size, so that the bubble generator generates bubbles of the first bubble particle size.

[0042] For example, for each candidate flocculant type, the floc particle size change curve at each moment in the floc particle size change graph is selected, the floc particle size corresponding to the peak value of the particle number in all floc particle size change curves is selected, and then the floc particle size corresponding to the peak value of the particle number of the floc particle size change curve at all moments is averaged to obtain the average floc particle size of the candidate flocculant type. The function expression of the average floc particle size is as follows: (1); in, represents the average floc particle size, It represents the floc particle size corresponding to the peak value of the number of particles in the floc particle size change curve at time k in the floc particle size change diagram. Represents the total number of floc size change curves in the floc size change graph.

[0043] For each candidate flocculant type, all flocculant particle size change curves in its flocculant particle size change diagram are selected, and the flocculant particle size corresponding to the maximum number of particles in all flocculant particle size change curves is selected. Based on the flocculant particle size corresponding to the maximum number of particles and the initial flocculant particle size of the corresponding candidate flocculant type, the flocculation efficiency of the candidate flocculant type is calculated. The function expression of flocculation efficiency is as follows: 00% (2); in, represents the flocculation efficiency, It indicates the floc particle size corresponding to the maximum number of particles in all floc particle size change curves in the floc particle size change diagram. The initial flocculent particle size of the corresponding candidate flocculants, Indicates the time length corresponding to the floc particle size change curve from the flocculation start time to the maximum particle number in the floc particle size change graph. In the embodiment of the present invention, the time when the flocs begin to form can be used as the flocculation start time.

[0044] After calculating the flocculation efficiency of each candidate flocculant type, select the candidate flocculant type corresponding to the highest flocculation efficiency as the target flocculant type. Then, based on the flocculation efficiency of the target flocculant type and the volume of the wastewater to be treated, calculate the first dosage of the flocculant corresponding to the target flocculant type. The function expression of the first dosage is as follows: (3); in, Indicates the first dosage, represents the volume of wastewater to be treated, Indicates the expected purification time. represents the initial particle concentration of the wastewater to be treated, Indicates the expected particle concentration of the wastewater to be treated. Initial particle concentration of the wastewater to be treated The expected particle concentration of the wastewater to be treated is obtained by calculating the sum of the concentrations of all oil droplet sizes in the first oil droplet size distribution information and the sum of the concentrations of all suspended matter sizes in the first suspended matter size distribution information. , Expected purification time It can be predefined and is not specifically limited in the embodiment of the present invention.

[0045] The embodiments of the present invention take into account that the molecular weight and charge density of the flocculant will significantly affect the adhesion of the bubbles on the surface of the flocculant, and the strength of the adhesion directly determines the degree of combination between the bubbles and the flocculant. Too strong or too weak will lead to a decrease in the efficiency of oil droplet demulsification. For example, a high molecular weight flocculant is easy to form a large volume and loose structure of flocculants, and the surface charge density is relatively low. In this case, the adhesion between the bubbles and the flocculants is weak, and the bubbles are easy to fall off from the surface of the flocculants, resulting in incomplete demulsification of the oil droplets; the flocculants formed by low molecular weight flocculants are small in volume, compact in structure, high in surface charge density, strong in bubble adhesion, and high in oil droplet demulsification efficiency, but too high in surface charge density will cause bubbles to excessively gather on the surface of the flocculants, forming a dense bubble layer, which hinders the full contact between the oil droplets and the flocculant, and reduces the demulsification efficiency. At the same time, the bubble particle size will also significantly affect the adhesion of the bubbles on the surface of the flocculants; larger bubbles usually have stronger adhesion due to the larger contact area with the liquid surface or solid particles. Based on this, the embodiment of the present invention establishes a matching relationship between the charge density of the flocculant and the bubble particle size through the indicator of the bubble attachment degree (the bubble attachment degree is used to describe the degree of adhesion (adhesion) between the bubbles and the surface of the flocculent body), and finds an optimal matching point to ensure that the combination of a flocculant with a specific charge density and bubbles with a specific bubble particle size can produce the best flocculation effect, thereby improving the wastewater purification efficiency and purification effect.

[0046] Specifically, a high-speed camera is used to capture an image sequence of the flocculation and demulsification process of oil droplets in an oily wastewater sample under the action of flocculants with different charge densities and bubbles of standard particle size, and the image sequence is identified by an image recognition tool to determine the image frames of flocculation initiation and demulsification under the action of flocculants with different charge densities. The demulsification time of the corresponding flocculant is obtained based on the difference between the corresponding timestamps of the image frame of flocculation initiation and the corresponding timestamps of the image frame of demulsification, and the demulsification time of the flocculants with different charge densities is recorded.

[0047] It should be understood that demulsification time refers to the time required for emulsion to form and decompose. Since the stronger bubble adhesion makes the bubbles and the liquid surface or solid particles more tightly bonded, the emulsion is more stable and not easy to decompose, and the demulsification time will be relatively long. Therefore, the greater the bubble adhesion, the longer the demulsification time.

[0048] Then, according to the charge density and demulsification time of different flocculants recorded above, the correlation function f( ).

[0049] Then, according to the correlation function f( ), calculate the demulsification time under the charge density of the flocculant corresponding to the target flocculant type, and combine the average flocculent particle size of the flocculant corresponding to the target flocculant type to calculate the target bubble attachment degree under the target flocculant type. The function expression of the target bubble attachment degree is as follows: (4); in, Indicates the target bubble adhesion degree, represents the charge density, Represents the charge density The demulsification time is Indicates the set particle size influence constant, Indicates the average flocculant particle size of the target flocculant type. , represents the weight coefficient, + =1, which describes the relationship between bubble adhesion, charge density and flocculent particle size.

[0050] Finally, according to the target bubble attachment degree, by querying the first mapping table between the bubble attachment degree and the bubble particle size established in advance, the bubble particle size corresponding to the target bubble attachment degree is obtained as the first bubble particle size. Similarly, the embodiment of the present invention also pre-establishes a second mapping table between different bubble particle sizes and pressure parameters. By querying the second mapping table, the pressure parameter corresponding to the first bubble particle size can be obtained. The bubble generator is configured to the pressure parameter, and the bubble generator can generate bubbles of the first bubble particle size.

[0051] The embodiment of the present invention associates the matching relationship between flocculant charge density and bubble particle size through the bubble adhesion degree, and can find the bubble particle size that best matches the target flocculant type, thereby ensuring the best flocculation effect and improving wastewater purification efficiency and purification effect.

[0052] In an optional embodiment, the method further includes: Real-time acquisition of a second near-infrared absorption spectrum during the purification process of the wastewater to be treated; Real-time acquisition of second particle size distribution information during the purification process of the wastewater to be treated; According to the second near-infrared absorption spectrum and the first corresponding relationship between the oil droplet diameter and the wavelength absorption peak, the particle diameter matching the oil droplet diameter and the concentration information of the corresponding particle diameter are screened out from the second particle diameter distribution information to obtain the second oil droplet diameter distribution information in the wastewater purification process to be treated; In an embodiment of the present invention, the second near-infrared absorption spectrum and the second particle size distribution information in the wastewater purification process to be treated can be obtained at a preset time interval. It should be noted that the process of obtaining the second oil droplet particle size distribution information can refer to the process of obtaining the first oil droplet particle size distribution information, which will not be repeated here.

[0053] The bubbles in the wastewater to be treated are monitored in real time by a high-speed camera, and the bubble images monitored in real time are recognized to obtain the bubble particle size distribution information in the wastewater to be treated; The embodiment of the present invention uses a high-speed camera to monitor the bubbles floating up in the flocculation stage of the wastewater purification process in real time, obtains continuous frame images within a set time granularity taken by the high-speed camera at preset time intervals, and extracts a frame image with the highest bubble edge clarity from the continuous frame images as a bubble image.

[0054] According to the second oil droplet particle size distribution information and the bubble particle size distribution information, a purification adjustment parameter is obtained through a pre-established second deep neural network model; wherein the purification adjustment parameter includes a second dosage of a flocculant corresponding to a target flocculant type and a second bubble particle size; Specifically, the purification adjustment parameters are obtained by using a pre-established second deep neural network model according to the second oil droplet size distribution information and the bubble size distribution information, including: Calculating the oil droplet size distribution width at the current moment according to the second oil droplet size distribution information at the current moment; When the oil droplet particle size distribution width is greater than the preset alarm particle size range, the purification adjustment parameters are obtained according to the second oil droplet particle size distribution information at the current moment and the bubble particle size distribution information at the current moment through a pre-established second deep neural network model.

[0055] Specifically, the purification adjustment parameters are obtained by using a pre-established second deep neural network model according to the second oil droplet size distribution information at the current moment and the bubble size distribution information at the current moment, including: Constructing an oil droplet size distribution feature vector according to the second oil droplet size distribution information at the current moment; According to the bubble particle size distribution information at the current moment, a bubble particle size distribution feature vector is constructed; Inputting the oil droplet size distribution feature vector and the bubble size distribution feature vector into the second deep neural network model to obtain a predicted bubble adhesion degree; According to the predicted bubble adhesion degree and the target bubble adhesion degree, a second bubble particle size and a second dosage are calculated.

[0056] Considering that during the wastewater purification process, the particle size of the suspended solids remains basically unchanged during the flocculation process, while the particle size of the oil pollution changes with the flocculation time during the flocculation process. Therefore, the embodiment of the present invention mainly considers the changes in the particle size of the oil droplets and the particle size of the bubbles during the flocculation process to dynamically adjust the dosage of the flocculant and the bubble particle size in the wastewater to be treated.

[0057] For example, by periodically detecting the second oil droplet size distribution information of the wastewater to be treated during the purification process, the maximum oil droplet size and the minimum oil droplet size are obtained from the second oil droplet size distribution information detected at the current moment, and the oil droplet size distribution width at the current moment is determined with the maximum oil droplet size as the upper limit and the minimum oil droplet size as the lower limit. When the oil droplet size distribution width is greater than the preset alarm particle size range, it is determined to dynamically adjust the flocculant dosage and bubble particle size; otherwise, the current flocculant dosage and bubble particle size are maintained. It should be noted that the embodiment of the present invention does not specifically limit the value of the alarm particle size range. For example, the alarm particle size range can be determined based on a preset error range centered on the oil droplet size distribution width corresponding to the first oil droplet size distribution information, such as the alarm particle size range = , represents the lower limit value of the oil droplet size distribution width corresponding to the first oil droplet size distribution information, represents the upper limit value of the oil droplet size distribution width corresponding to the first oil droplet size distribution information, Indicates the preset error value.

[0058] When it is determined that the dosage of flocculant and bubble size need to be dynamically adjusted, the second oil droplet size distribution information at the current moment is used to construct the oil droplet size distribution feature vector; for example, the oil droplet size distribution feature vector , represents the i-th oil droplet size in the second oil droplet size distribution information, Indicates the concentration of the i-th oil droplet size in the second oil droplet size distribution information. Similarly, based on the bubble size distribution information at the current moment, a bubble size distribution feature vector is constructed.

[0059] The above-mentioned oil droplet size distribution feature vector and bubble size distribution feature vector are input into the pre-trained second deep neural network model to predict the bubble adhesion degree, and obtain the predicted bubble adhesion degree; it should be noted that the second deep neural network model can be constructed using a long short-term memory neural network LSTM, which is not specifically limited in the embodiments of the present invention. By collecting the oil droplet size distribution information, bubble size distribution information and corresponding bubble adhesion degree (specific calculations can be found in the above formula (4)) of different wastewater samples in the flocculation process under the action of bubbles of different bubble particle sizes as training samples, the long short-term memory neural network LSTM is trained to obtain the second deep neural network model. Among them, the training process of the long short-term memory neural network LSTM belongs to the prior art and will not be described in detail here.

[0060] By querying a pre-established first mapping table between the bubble adhesion degree and the bubble particle size, the bubble particle size corresponding to the predicted bubble adhesion degree is obtained as the second bubble particle size.

[0061] Then, the absolute value of the difference between the predicted bubble adhesion degree and the target bubble adhesion degree is calculated. The second dosage of the target flocculant type corresponding to the flocculant ,in, represents the absolute value of the difference between the predicted bubble attachment degree and the target bubble attachment degree, Indicates the set influence factor. When the predicted bubble attachment degree is greater than or equal to the target bubble attachment degree, ; When the predicted bubble attachment degree is less than the target bubble attachment degree, . Indicates a preset constant.

[0062] The flocculant dosing pump is adjusted to add a corresponding flocculant dosage to the wastewater to be treated according to the second dosage, and the bubble generator is controlled to input bubbles of the second bubble size into the wastewater to be treated, so as to purify and optimize the wastewater to be treated.

[0063] After calculating the flocculant dosage and bubble particle size that need to be adjusted, the system controls the flocculant dosage pump to adjust from the first dosage to the second dosage, and at the same time controls the bubble particle size generated by the bubble generator to adjust from the first bubble particle size to the second bubble particle size, so as to periodically and dynamically adjust the flocculant dosage and bubble particle size according to the actual flocculation effect of the wastewater purification process, thereby further improving the wastewater purification effect.

[0064] See also Figure 4 , Figure 4 The embodiment of the present invention provides a structural block diagram of a power transmission and transformation project wastewater purification treatment system based on a neural network, and the power transmission and transformation project wastewater purification treatment system based on a neural network includes: The first particle size detection module 11 is used to detect the first particle size distribution information of pollutants in the wastewater to be treated in the power transmission and transformation project through a laser particle size detector; A first near-infrared absorption spectrum acquisition module 12, used to acquire a first near-infrared absorption spectrum of the wastewater to be treated; A first particle size identification module 13 is used to identify oil droplets and suspended matter for pollutants corresponding to different particle sizes in the first particle size distribution information according to the first near-infrared absorption spectrum and a pre-established first correspondence between oil droplet particle size and wavelength absorption peak, and a second correspondence between suspended matter particle size and wavelength absorption peak, so as to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; A purification parameter determination module 14 is used to determine purification parameters according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information through a pre-established first deep neural network model; wherein the purification parameters include a target flocculant type, a first dosage, and a first bubble particle size; The first control module 15 is used to control the flocculant dosing pump to add the corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage, and configure the pressure parameter of the bubble generator to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated to purify the wastewater to be treated.

[0065] In an optional embodiment, the system further includes: A second near-infrared absorption spectrum acquisition module, used for acquiring in real time the second near-infrared absorption spectrum in the purification process of the wastewater to be treated; A second particle size detection module is used to obtain second particle size distribution information in real time during the purification process of the wastewater to be treated; A second particle size identification module is used to screen out the particle size matching the oil droplet particle size and the concentration information of the corresponding particle size from the second particle size distribution information according to the second near-infrared absorption spectrum and the first corresponding relationship between the oil droplet particle size and the wavelength absorption peak, so as to obtain the second oil droplet particle size distribution information in the wastewater purification process to be treated; A bubble monitoring module is used to monitor the bubbles in the wastewater to be treated in real time through a high-speed camera, and to perform image recognition on the bubble images monitored in real time to obtain the bubble particle size distribution information in the wastewater to be treated in the wastewater to be treated; A purification adjustment parameter determination module, used to obtain purification adjustment parameters through a pre-established second deep neural network model according to the second oil droplet particle size distribution information and the bubble particle size distribution information; wherein the purification adjustment parameters include a second dosage of flocculant corresponding to the target flocculant type and a second bubble particle size; The second control module is used to adjust the dosage of the flocculant dosing pump to add the corresponding flocculant to the wastewater to be treated according to the second dosage, and control the bubble generator to input bubbles of the second bubble size into the wastewater to be treated, so as to purify and optimize the wastewater to be treated.

[0066] In an optional embodiment, the system further includes: A third near-infrared absorption spectrum acquisition module, used to acquire third near-infrared absorption spectra of a plurality of oily wastewater samples containing oil droplet particle sizes of different sizes; A first corresponding relationship establishing module, used to determine the wavelength absorption peaks corresponding to different oil droplet particle sizes according to the third near-infrared absorption spectra of different oil droplet particle sizes, and establish a first corresponding relationship between the oil droplet particle size and the wavelength absorption peak; A fourth near-infrared absorption spectrum acquisition module is used to acquire fourth near-infrared absorption spectra of a plurality of suspended matter-containing wastewater samples having different suspended matter particle sizes; The second corresponding relationship establishing module is used to determine the wavelength absorption peaks corresponding to different suspended matter particle sizes according to the fourth near-infrared absorption spectra of different suspended matter particle sizes, and establish a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak.

[0067] In an optional embodiment, the first particle size identification module 13 includes: a target oil droplet diameter determination unit, configured to obtain the target oil droplet diameter in the wastewater to be treated according to the first near-infrared absorption spectrum and a first corresponding relationship between the oil droplet diameter and the wavelength absorption peak; an oil droplet size distribution information acquisition unit, configured to select a particle size matching the target oil droplet size and concentration information of the corresponding particle size from the first particle size distribution information, and obtain the first oil droplet size distribution information; a target suspended matter particle size determination unit, configured to obtain the target suspended matter particle size in the wastewater to be treated according to the first near-infrared absorption spectrum and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak; The suspended matter particle size distribution information acquisition unit is used to select the particle size matching the target suspended matter particle size and the concentration information of the corresponding particle size from the first particle size distribution information to obtain the first suspended matter particle size distribution information.

[0068] In an optional embodiment, the purification parameter determination module 14 includes: a flocculant type screening unit, configured to screen out at least one candidate flocculant type from a flocculant database according to the first oil droplet size distribution information and the first suspended matter size distribution information; A wastewater characteristic vector construction unit, configured to construct a wastewater characteristic vector according to the first oil droplet particle size distribution information, the first suspended matter particle size distribution information, and the current pH value and current water temperature of the wastewater to be treated; A flocculant characteristic vector construction unit, used to construct a flocculant characteristic vector of a corresponding candidate flocculant type according to a molecular weight and a charge density of a flocculant corresponding to at least one candidate flocculant type; A flocculant particle size change diagram prediction unit, used for inputting the wastewater feature vector and the flocculant feature vector corresponding to each of the candidate flocculant types into the first deep neural network model to obtain a flocculant particle size change diagram of the flocculant corresponding to each of the candidate flocculant types in the wastewater to be treated; The parameter determination unit is used to determine the purification parameters according to the flocculant particle size change diagram corresponding to each of the candidate flocculant types.

[0069] In an optional embodiment, the parameter determination unit includes: A flocculation parameter calculation subunit, for calculating the flocculation efficiency and average flocculant particle size of each candidate flocculant type according to the flocculant particle size change diagram corresponding to each candidate flocculant type; A target flocculant type determination subunit is used to select a candidate flocculant type corresponding to the highest flocculation efficiency as the target flocculant type; A first dosage calculation subunit, used for calculating a first dosage of a flocculant corresponding to the target flocculant type according to the flocculation efficiency of the target flocculant type; The first bubble particle size calculation subunit is used to calculate the first bubble particle size according to the charge density and average flocculent particle size of the flocculant corresponding to the target flocculant type.

[0070] In an optional embodiment, the first bubble diameter calculation subunit includes: A bubble attachment degree calculation subunit, used to calculate a target bubble attachment degree under the target flocculant type according to the charge density and average flocculant particle size of the flocculant corresponding to the target flocculant type; A query subunit, configured to obtain the first bubble particle size according to the target bubble attachment degree and a pre-established first mapping table between the bubble attachment degree and the bubble particle size; The pressure parameter of the bubble generator is configured as a pressure parameter corresponding to the first bubble particle size, so that the bubble generator generates bubbles of the first bubble particle size.

[0071] In an optional embodiment, the purification adjustment parameter determination module includes: An oil droplet size distribution width calculation unit, used to calculate the oil droplet size distribution width at the current moment according to the second oil droplet size distribution information at the current moment; The purification adjustment parameter prediction unit is used to obtain the purification adjustment parameter through a pre-established second deep neural network model according to the second oil droplet size distribution information at the current moment and the bubble size distribution information at the current moment when the oil droplet size distribution width is greater than the preset alarm particle size range.

[0072] In an optional embodiment, the purification adjustment parameter prediction unit includes: An oil droplet size distribution feature vector construction subunit is used to construct an oil droplet size distribution feature vector according to the second oil droplet size distribution information at the current moment; A bubble particle size distribution feature vector construction subunit is used to construct a bubble particle size distribution feature vector according to the bubble particle size distribution information at the current moment; a bubble adhesion degree prediction subunit, used for inputting the oil droplet size distribution feature vector and the bubble size distribution feature vector into the second deep neural network model to obtain a predicted bubble adhesion degree; The parameter calculation subunit is used to calculate the second bubble particle size and the second dosage according to the predicted bubble adhesion degree and the target bubble adhesion degree.

[0073] It should be noted that the working process of each module in the power transmission and transformation project wastewater purification treatment system based on neural network described in the embodiment of the present invention can refer to the working process of the power transmission and transformation project wastewater purification treatment method based on neural network described in the above embodiment, and the technical effect achieved is also the same as the power transmission and transformation project wastewater purification treatment method based on neural network described in the above embodiment, which will not be repeated here.

[0074] It should be noted that the system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, in the drawings of the system embodiment provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art may understand and implement it without paying any creative effort.

[0075] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, many improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for purifying wastewater from power transmission and transformation projects based on neural networks, characterized in that: include: Detecting the first particle size distribution information of pollutants in the wastewater to be treated in the power transmission and transformation project by using a laser particle size detector; Acquiring a first near-infrared absorption spectrum of the wastewater to be treated; According to the first near-infrared absorption spectrum and the pre-established first correspondence between the oil droplet particle size and the wavelength absorption peak, and the pre-established second correspondence between the suspended matter particle size and the wavelength absorption peak, the pollutants corresponding to different particle sizes in the first particle size distribution information are identified as oil droplets and suspended matter, to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; According to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, purification parameters are determined by a pre-established first deep neural network model; wherein the purification parameters include a target flocculant type, a first dosage, and a first bubble particle size; The flocculant dosing pump is controlled to add the corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage, and the pressure parameter of the bubble generator is configured to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated to purify the wastewater to be treated.

2. The method for purifying wastewater from power transmission and transformation projects based on a neural network as claimed in claim 1, characterized in that: The method further comprises: Real-time acquisition of a second near-infrared absorption spectrum during the purification process of the wastewater to be treated; Real-time acquisition of second particle size distribution information during the purification process of the wastewater to be treated; According to the second near-infrared absorption spectrum and the first corresponding relationship between the oil droplet diameter and the wavelength absorption peak, the particle diameter matching the oil droplet diameter and the concentration information of the corresponding particle diameter are screened out from the second particle diameter distribution information to obtain the second oil droplet diameter distribution information in the wastewater purification process to be treated; The bubbles in the wastewater to be treated are monitored in real time by a high-speed camera, and the bubble images monitored in real time are recognized to obtain the bubble particle size distribution information in the wastewater to be treated; According to the second oil droplet particle size distribution information and the bubble particle size distribution information, a purification adjustment parameter is obtained through a pre-established second deep neural network model; wherein the purification adjustment parameter includes a second dosage of a flocculant corresponding to a target flocculant type and a second bubble particle size; The flocculant dosing pump is adjusted to add a corresponding flocculant dosage to the wastewater to be treated according to the second dosage, and the bubble generator is controlled to input bubbles of the second bubble size into the wastewater to be treated, so as to purify and optimize the wastewater to be treated.

3. The method for purifying wastewater from power transmission and transformation projects based on neural network according to claim 1, characterized in that: The method further comprises: Obtaining third near-infrared absorption spectra of multiple oily wastewater samples containing different oil droplet sizes; Determine the wavelength absorption peaks corresponding to the different oil droplet sizes according to the third near-infrared absorption spectra of the different oil droplet sizes, and establish a first corresponding relationship between the oil droplet size and the wavelength absorption peak; obtaining fourth near infrared absorption spectra of a plurality of suspended matter wastewater samples having different suspended matter particle sizes; According to the fourth near-infrared absorption spectra of different suspended matter particle sizes, wavelength absorption peaks corresponding to different suspended matter particle sizes are determined, and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak is established.

4. The method for purifying wastewater from power transmission and transformation projects based on a neural network as claimed in claim 3, characterized in that: According to the first near-infrared absorption spectrum and the pre-established first correspondence between the oil droplet particle size and the wavelength absorption peak, and the pre-established second correspondence between the suspended matter particle size and the wavelength absorption peak, the pollutants corresponding to different particle sizes in the first particle size distribution information are identified with oil droplets and suspended matter to obtain the first oil droplet particle size distribution information and the first suspended matter particle size distribution information, including: According to the first near-infrared absorption spectrum and the first corresponding relationship between the oil droplet diameter and the wavelength absorption peak, the target oil droplet diameter in the wastewater to be treated is obtained; Screening out the particle size matching the target oil droplet particle size and the concentration information of the corresponding particle size from the first particle size distribution information to obtain the first oil droplet particle size distribution information; According to the first near-infrared absorption spectrum and a second corresponding relationship between the suspended matter particle size and the wavelength absorption peak, the target suspended matter particle size in the wastewater to be treated is obtained; The particle size matching the target suspended matter particle size and the concentration information of the corresponding particle size are screened out from the first particle size distribution information to obtain the first suspended matter particle size distribution information.

5. The method for purifying wastewater from power transmission and transformation projects based on neural network according to claim 1, characterized in that: The step of determining purification parameters according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information by using a pre-established first deep neural network model includes: Screening out at least one candidate flocculant type from a flocculant database according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information; Constructing a wastewater feature vector according to the first oil droplet particle size distribution information, the first suspended matter particle size distribution information, and the current pH value and current water temperature of the wastewater to be treated; According to the molecular weight and charge density of the flocculant corresponding to at least one candidate flocculant type, construct a flocculant feature vector of the corresponding candidate flocculant type; Inputting the wastewater feature vector and the flocculant feature vectors corresponding to each of the candidate flocculant types into the first deep neural network model to obtain a flocculant particle size change graph of the flocculant corresponding to each of the candidate flocculant types in the wastewater to be treated; The purification parameters are determined according to the flocculent particle size variation diagram corresponding to each of the candidate flocculant types.

6. The method for purifying wastewater from power transmission and transformation projects based on a neural network as claimed in claim 5, characterized in that: The step of determining the purification parameters according to the flocculant particle size variation diagram corresponding to each candidate flocculant type includes: Calculating the flocculation efficiency and average flocculent particle size of each candidate flocculant type according to the flocculent particle size change diagram corresponding to each candidate flocculant type; Select the candidate flocculant type corresponding to the highest flocculation efficiency as the target flocculant type; Calculating a first dosage of a flocculant corresponding to the target flocculant type according to the flocculation efficiency of the target flocculant type; The first bubble particle size is calculated according to the charge density of the flocculant corresponding to the target flocculant type and the average flocculent particle size.

7. The method for purifying wastewater from power transmission and transformation projects based on a neural network as claimed in claim 6, characterized in that: The calculating the first bubble particle size according to the charge density of the flocculant corresponding to the target flocculant type includes: Calculating a target bubble attachment degree under the target flocculant type according to the charge density and average flocculent particle size of the flocculant corresponding to the target flocculant type; The first bubble particle size is obtained according to the target bubble adhesion degree and a pre-established first mapping table between the bubble adhesion degree and the bubble particle size.

8. The method for purifying wastewater from power transmission and transformation projects based on a neural network as claimed in claim 2, characterized in that: The step of obtaining purification adjustment parameters according to the second oil droplet size distribution information and the bubble size distribution information by using a pre-established second deep neural network model includes: Calculating the oil droplet size distribution width at the current moment according to the second oil droplet size distribution information at the current moment; When the oil droplet particle size distribution width is greater than the preset alarm particle size range, the purification adjustment parameters are obtained according to the second oil droplet particle size distribution information at the current moment and the bubble particle size distribution information at the current moment through a pre-established second deep neural network model.

9. The method for purifying wastewater from power transmission and transformation projects based on a neural network as claimed in claim 8, characterized in that: The purification adjustment parameters are obtained by using a pre-established second deep neural network model according to the second oil droplet particle size distribution information at the current moment and the bubble particle size distribution information at the current moment, including: Constructing an oil droplet size distribution feature vector according to the second oil droplet size distribution information at the current moment; According to the bubble particle size distribution information at the current moment, a bubble particle size distribution feature vector is constructed; Inputting the oil droplet size distribution feature vector and the bubble size distribution feature vector into the second deep neural network model to obtain a predicted bubble adhesion degree; According to the predicted bubble adhesion degree and the target bubble adhesion degree, a second bubble particle size and a second dosage are calculated.

10. A power transmission and transformation engineering wastewater purification system based on neural network, characterized in that: include: A first particle size detection module is used to detect first particle size distribution information of pollutants in wastewater to be treated in a power transmission and transformation project through a laser particle size detector; A first near-infrared absorption spectrum acquisition module, used to acquire a first near-infrared absorption spectrum of the wastewater to be treated; a first particle size identification module, for identifying oil droplets and suspended matter for pollutants corresponding to different particle sizes in the first particle size distribution information according to the first near-infrared absorption spectrum and a pre-established first correspondence between oil droplet particle size and wavelength absorption peak, and a pre-established second correspondence between suspended matter particle size and wavelength absorption peak, to obtain first oil droplet particle size distribution information and first suspended matter particle size distribution information; A purification parameter determination module, used to determine purification parameters according to the first oil droplet particle size distribution information and the first suspended matter particle size distribution information through a pre-established first deep neural network model; wherein the purification parameters include a target flocculant type, a first dosage, and a first bubble particle size; The first control module is used to control the flocculant dosing pump to add the corresponding flocculant to the wastewater to be treated according to the target flocculant type and the first dosage, and configure the pressure parameter of the bubble generator to the pressure parameter corresponding to the first bubble particle size, so that the bubble generator inputs bubbles of the first bubble particle size into the wastewater to be treated to purify the wastewater to be treated.

Images