Fabricated pit three-dimensional construction and point cloud data sewage detection method and device

Abstract

The embodiment of the invention provides an assembly type pit three-dimensional construction and point cloud data sewage detection method and device. The method is applied to the technical field of intelligent water quality evaluation, and comprises the following steps: collecting actual measurement data and laser point cloud data of a winery sewage body, and performing data processing to obtain processed sewage body pollutant data and laser point cloud data; constructing a sewage water body three-dimensional visual model according to the laser point cloud data to obtain a water body three-dimensional model; constructing a training set and a test set according to the sewage water pollutant data, selecting the training set, and training and constructing a sewage evaluation model based on an AdaBoost strategy; evaluating the winery sewage body according to the sewage evaluation model, and calculating to obtain the winery sewage pollutant concentration based on the evaluation data result and the water body three-dimensional model. In this way, high-dimensional data can be effectively processed, the volume of the water body can be effectively calculated through the three-dimensional point cloud data, and the accuracy and robustness of concentration evaluation are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent water quality assessment, and in particular to a method and device for three-dimensional construction of assembled pits and sewage detection using point cloud data. Background Art

[0002] Alcohol is widely used in chemical, food, daily chemical, medical and health fields. It is also a wine base, extractant, solvent, detergent and surfactant. The production of alcohol is also accompanied by the discharge of sewage.

[0003] Winery wastewater is high-concentration, high-temperature, high-suspended organic wastewater. The waste residue in the wastewater contains crushed fermentation raw materials. Predicting water quality changes is one of the challenging problems. The main water quality monitoring methods are mainly used for short-term water quality concentration assessment with strong trends and small fluctuations, but it is difficult to implement for large-scale training samples and has certain limitations. Therefore, there is an urgent need for an assessment method that can be used for large-scale winery wastewater water quality concentration. Summary of the invention

[0004] The present invention provides a method and device for sewage detection based on the three-dimensional construction of an assembled pit and point cloud data. The AdaBoost algorithm can effectively process high-dimensional data, improve the accuracy and robustness of concentration assessment, and solve the technical problem that the prior art cannot perform concentration assessment on large-scale water quality data.

[0005] According to a first aspect of the present disclosure, a method for three-dimensional construction of an assembled pit and sewage detection using point cloud data is provided, comprising the following steps:

[0006] Collect measured data of winery sewage water bodies and laser point cloud data, and process the data to obtain pollutant data of treated sewage water bodies and laser point cloud data;

[0007] Constructing a three-dimensional visualization model of the sewage water body according to the laser point cloud data to obtain a three-dimensional model of the water body;

[0008] Constructing a training set and a test set according to the sewage water pollutant data, selecting a training set, and constructing a sewage assessment model based on AdaBoost strategy training;

[0009] Evaluate the winery wastewater water body according to the wastewater evaluation model, output the evaluation data results, and calculate the winery wastewater pollutant concentration based on the evaluation data results and the three-dimensional model of the water body;

[0010] The process of constructing a training set and a test set according to the sewage water pollutant data, selecting a training set, and constructing a sewage assessment model based on AdaBoost strategy training is as follows:

[0011] Obtain training samples based on the training set and initialize a training sample weight matrix;

[0012] A sub-learning method is adopted and a sub-learner is obtained by training based on the training set after the weights are initialized;

[0013] Calculating the training error and weight matrix of the sub-learner;

[0014] The weight matrix of the training sample is obtained by continuously iterating the training and adjusting according to the training error and the weight matrix;

[0015] Integrate the weight matrices of sub-learners based on the integration strategy and output a strong learner;

[0016] The process of evaluating the winery sewage water body according to the sewage evaluation model, outputting the evaluation data results, and calculating the winery sewage pollutant concentration based on the evaluation data results and the water body three-dimensional model is as follows:

[0017] The total nitrogen value data of the sewage is obtained by evaluating the sewage evaluation model, and the volume of the sewage water body is obtained by calculating the three-dimensional model of the water body;

[0018] The total nitrogen concentration value in the winery wastewater is calculated based on the total nitrogen value data of the wastewater and the volume of the wastewater;

[0019] The sewage water volume calculation process is:

[0020] Slicing the three-dimensional water body model horizontally at equal intervals to obtain a plurality of slice polygons;

[0021] The α-shape method is used to iteratively find the threshold of each point cloud slice polygon, and the area of ​​the outer contour of the slice polygon is obtained based on the discretized Green's formula under this threshold;

[0022] Calculating the straight-line distance between the slice polygons, and obtaining the volume of the three-dimensional model according to the straight-line distance and the area calculation;

[0023] The volume calculation process of the three-dimensional model is as follows:

[0024]

[0025] Among them, Δ is the straight-line distance between two adjacent slices, n′ is the number of slices, and d i′ is the water body fault coefficient;

[0026] The calculation process of the water body fault coefficient is as follows:

[0027] Obtain the plane boundary of the water body slice breakpoint data, and use the plane α-shape edge extraction algorithm to obtain the convex hull boundary of the target fault breakpoint data on the current time slice;

[0028] The polygonal paranoid method with inclination constraints is used to calculate the convex hull expansion polygon and obtain the spatial boundary of the target fault of adjacent time slices;

[0029] Search for the breakpoint data within the convex hull expansion polygon on the adjacent time slices, and determine whether it is an empty set. If so, output the plane shape of the target water body fault. Otherwise, determine whether the breakpoint data belongs to the water body fault. The process of determining whether the breakpoint data belongs to the water body fault is as follows: for the breakpoint data within the convex hull expansion polygon, set the target fault overlap threshold. If the overlap is greater than the threshold, it is determined to be a non-target fault. If it is less than the threshold, it is determined to belong to the target fault.

[0030] If yes, the target water body fault plane shape is output; if no, the adjacent time slice is updated to the current time slice, and the breakpoint data and convex hull boundary are updated;

[0031] The current breakpoint data is merged into the target water body fault, the target water body fault space boundary is merged, the target fault plane morphology is obtained, and the Euclidean distance between the center points of the adjacent target fault spaces is used as the water body fault coefficient.

[0032] According to the above aspects and any possible implementation, an implementation is further provided, wherein the process of collecting measured data of winery sewage water body and laser point cloud data, and performing data processing to obtain treated sewage water body pollutant data and laser point cloud data is as follows:

[0033] Several laser point cloud scanners are deployed in the sewage water body to collect laser point cloud data, so as to obtain sewage water body point cloud data; the sewage water body is collected at specific time intervals, and the total nitrogen content in the water body is measured by chromotropic acid spectrophotometry to obtain sewage water body pollutant data.

[0034] According to the above aspects and any possible implementation, an implementation is further provided, wherein a plurality of laser point cloud scanners are arranged in the sewage body to collect laser point cloud data, and the process of obtaining the sewage body point cloud data is as follows:

[0035] Using a light intensity regulation circuit and an A / D acquisition module to send an acquisition command to the laser scanner;

[0036] After receiving the command, the laser scanner sends the collected laser point cloud data to the DSP through the light intensity regulation circuit for buffering;

[0037] The cached laser point cloud data is sent to the data server to complete the sewage water body point cloud data collection process.

[0038] According to the above aspects and any possible implementation, an implementation is further provided, wherein the process of constructing the three-dimensional model of the water body is:

[0039] Convert the collected laser point cloud data into a coordinate system;

[0040] The laser point cloud data after coordinate conversion is imported into 3D Max for processing to construct the model framework, and then three-dimensional modeling is performed to obtain a three-dimensional model of the water body.

[0041] According to the above aspects and any possible implementation manner, an implementation manner is further provided, wherein the coordinate system conversion process is:

[0042] Calculating the center of the original point of the laser point cloud data, and calculating the center of the converted laser point cloud data;

[0043] A covariance matrix is ​​constructed based on the center of the original point of the laser point cloud data and the center of the converted laser point cloud data;

[0044] A multidimensional matrix is ​​constructed according to the covariance matrix, eigenvalues ​​and eigenvectors are calculated according to the multidimensional matrix, and the eigenvector corresponding to the maximum eigenvalue is selected to form a quaternion;

[0045] The laser point cloud rotation matrix and the translation matrix are obtained according to the quaternion calculation, and the rotation vector and the translation vector are calculated to determine the laser point cloud data after the coordinate transformation.

[0046] The aspects described above and any possible implementation methods further provide an implementation method, in which the construction process of the winery is as follows: multiple briefings are conducted during construction; the main body of the cellar pit is excavated by manual excavation or pneumatic pickaxe crushing within a range of 1m around the central column foundation; formwork is supported, concrete rear walls are poured, maintenance and high formwork removal, and wall thickness backfilling processes are carried out; and functional area construction is carried out.

[0047] According to the above aspects and any possible implementation, a further implementation is provided, wherein the concentration of the sewage pollutants is calculated as follows:

[0048]

[0049] Among them, t is the total nitrogen content data of the water body, c is the total nitrogen concentration of the water body, and V is the volume of the three-dimensional model of the water body.

[0050] According to a second aspect of the present disclosure, there is provided a sewage detection device for three-dimensional construction of assembled pits and point cloud data, including a water body data acquisition module, a water body three-dimensional model construction module, a sewage assessment model construction module, and a sewage pollutant concentration assessment module;

[0051] The water body data acquisition module is used to collect the measured data of the winery sewage water body and the laser point cloud data, and perform data processing to obtain the treated sewage water body pollutant data and the laser point cloud data;

[0052] The water body three-dimensional model construction module is used to construct a three-dimensional visualization model of the sewage water body according to the laser point cloud data to obtain a three-dimensional water body model;

[0053] The sewage assessment model construction module is used to construct a training set and a test set according to the sewage water body pollutant data, select a training set, and construct a sewage assessment model based on AdaBoost strategy training;

[0054] The sewage pollutant concentration assessment module is used to assess the winery sewage water body according to the sewage assessment model, output the assessment data results, and calculate the winery sewage pollutant concentration based on the three-dimensional model of the water body of the assessment data result set.

[0055] Compared with the prior art, the present invention has achieved the following beneficial technical effects:

[0056] (1) The three-dimensional construction of assembled pits and the sewage detection method and device of point cloud data proposed in the present invention can effectively process high-dimensional data through the AdaBoost algorithm, which is applicable to a large amount of water quality data. It does not require excessive feature selection or dimensionality reduction processing, and can solve the problem of uneven distribution of label values ​​that is difficult to solve by traditional machine learning algorithms, thereby improving evaluation efficiency and accuracy.

[0057] (2) Since the structures of sewage discharge channels of wineries are not the same, the present invention uses point cloud data to construct a three-dimensional model of the winery water body according to the actual situation and uses a plane slicing method to calculate the volume of the water body. It can more effectively calculate the concentration of sewage pollutants in the water body and has a high practical application value.

[0058] It should be understood that the contents described in the summary of the invention are not intended to limit the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] The above and other features, advantages and aspects of the embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0060] Figure 1 A schematic diagram of a sewage detection method for three-dimensional construction of an assembled pit and point cloud data according to an embodiment of the present disclosure is shown;

[0061] Figure 2 A schematic structural diagram of a sewage detection device for three-dimensional construction of an assembled pit and point cloud data according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0063] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] Please refer to Figure 1 As shown, this embodiment provides a method for three-dimensional construction of an assembled pit and sewage detection using point cloud data, including the following steps:

[0065] S101. Collect measured data of winery wastewater and laser point cloud data, and process the data.

[0066] Winery sewage is wastewater generated during the brewing process of wine, which is generally discharged through the sewage channel of the winery. The winery of this embodiment is built by the reverse construction method of the cellar. The specific process includes: when the cellar is reversed, the soil around the central column needs to be excavated. Therefore, it may cause damage to the structure during the construction process. This method innovatively uses mortise and tenon type easy-to-remove steel casings to protect the pile structure through the steel casing, and after the excavation of the earthwork is completed, the mortise and tenon steady state is destroyed by removing the pins to realize the turnover of the steel casing, while protecting the safety and stability of the structure, low-carbon green construction is achieved. When the cellar is reversed, because the cellar construction is synchronously constructed with the second structure and rough decoration, in order to avoid the inconvenience caused by cross-operation, this method innovatively uses prefabricated cellar backfill templates. In the drawing review stage, the strip foundation is optimized with the relevant units, the prefabricated template slot is set, and the prefabricated cellar backfill template is fixed. It can be realized that the outer wall of the cellar is grouting (waterproof mortar and anti-corrosion mortar grouting) without affecting the construction of functional areas such as the steamer area. A protective layer pier is set on the prefabricated pit backfill template to ensure that the thickness of the grouting (waterproof mortar and anti-corrosion mortar) meets the requirements of the specification. During the production process of prefabricated components, strict instructions are given to ensure that the size of the embedded slot is correct, the lifting lugs are pre-buried in advance, the maintenance of the prefabricated components is done well, and the strength is measured regularly to ensure that the strength meets the requirements during on-site implementation and does not affect the backfilling in the same period. Before prefabrication, the pit drawings are deepened to determine the backfill range between the steamer area and the pit, and the production is deepened according to the range. During prefabrication, according to the deepening drawings, ensure that the protective layer pads are produced according to the required thickness to ensure the thickness of the grouting material during the later grouting construction. Avoid deviations in the thickness of the waterproof and anti-corrosion mortar. Strict instructions are given when dismantling the upper and lower sections, and the whole process is supervised. The upper section is dismantled first, and the upper section is destroyed by prying the anchor nails to break the steady state, so as to achieve the separation of the upper and lower sections. Before the implementation of the mortise and tenon type easy-to-remove steel casing, the drawings of the workshop that needs to be constructed in reverse flow are deepened, the size of the middle pile is optimized, and the production is deepened to ensure that the on-site implementation can be turned around.

[0067] In this embodiment, a laser point cloud scanner is set in the sewage channel to collect laser point cloud data, and a DSP signal processing chip is used as the hardware for controlling the laser point cloud data collection. The DSP hardware is connected to a key control circuit, through which a laser point cloud data collection command is received, and a light intensity regulation circuit and an A / D acquisition module are used to send an acquisition command to the laser scanner. The laser scanner sends the collected laser point cloud data to the DSP through the light intensity regulation circuit. After caching the laser point cloud data, the DSP can send it to the database server through the peripheral interface, thereby realizing the laser point cloud data collection of the sewage water body.

[0068] After the laser point cloud data of the sewage water body is collected, the measured data of the winery sewage water body is collected. In this embodiment, a plurality of appropriate sampling points are determined in the water body, and sampling is performed at intervals of days to collect the collected water samples.

[0069] The total nitrogen content in the water sample is measured by chromotropic acid spectrophotometry. The water sample is digested under alkaline conditions with potassium persulfate as an oxidant. After high temperature and high pressure digestion, all nitrogen-containing compounds are converted into nitrate nitrogen. The remaining potassium persulfate is eliminated by adding excess sodium pyrosulfite. Nitrate-nitrogen reacts with chromotropic acid under strong acidic conditions to form a yellow compound. The absorbance value of the color developing solution is proportional to the total nitrogen content in the water sample.

[0070] After collecting the data, outliers are removed and missing values ​​are supplemented to complete the preprocessing process.

[0071] S102. Construct a three-dimensional visualization model of the sewage water body according to the laser point cloud data to obtain a three-dimensional model of the water body.

[0072] First, the collected water body point cloud data is transformed into a coordinate system. This embodiment uses the four-element method to transform the coordinates of the water body laser point cloud data. The specific process is:

[0073] Calculate the center of the original point of the laser point cloud data of the water body, specifically:

[0074]

[0075] Where z is the center of the original point of the laser point cloud data, n is the number of point cloud data, i = 1, 2, 3, ..., n, x i is the i-th point with a three-dimensional direction.

[0076] Then, calculate the center of the converted laser point cloud data as follows:

[0077]

[0078] Among them, z h is the center of the laser point cloud data after conversion, n′ is the number of converted laser point cloud data, y i is the i-th transformed point with 3D orientation.

[0079] Based on the original center point and the transformed center point of the laser point cloud data, the covariance matrix is ​​established, which is:

[0080]

[0081] Where R(c) is the covariance matrix and T is the normal distribution.

[0082] Based on the covariance matrix, a 4*4 matrix is ​​established to obtain:

[0083]

[0084] Where R(t) is a 4*4 matrix, tr(R(c)) is the trace of R(c), I3 is the identity matrix of dimension 3, and Δ is a matrix variable.

[0085] By calculating the eigenvalues ​​and eigenvectors of formula (4) and selecting the eigenvector corresponding to the maximum eigenvalue, the unit quaternion is obtained, specifically [q0q1q2,q3].

[0086] Finally, the laser point cloud rotation matrix and the translation matrix of the rotation matrix are calculated, specifically:

[0087] q T =z h -z·R(q R ) (5)

[0088] Among them, R(q R ) is the laser point cloud rotation matrix, q T R(q R ) is the translation matrix of .

[0089] After calculating the rotation and translation vectors of the laser point cloud data of the sewage water body, the coordinate transformation is completed.

[0090] The laser point cloud data after coordinate conversion is imported into 3DMax software to establish a model framework and construct a three-dimensional model of the water body.

[0091] S103, constructing a training set and a test set according to the measured water body data, selecting a training set, and constructing a sewage assessment model based on AdaBoost strategy training.

[0092] In order to construct a sewage assessment model, this embodiment uses the AdaBoost classification method to perform multi-classification modeling on the data set and establishes a regression model, which specifically includes an integration step and an iteration step.

[0093] Specifically, first initialize the training sample weight w t , the weight of each sample in each training is the same, which is 1 / N;

[0094] Then use the weighted sample set to train the data and get the sub-learner z t (x k );

[0095] Then the sub-learner z is calculated t (x k )’s training error e t , specifically:

[0096]

[0097] Where N is the total number of samples, x kis the kth sample, b k is the label value of the kth sample, w t is the weight matrix of the sample set after the tth iteration, Γ(b k ≠z t (x k )) is the sample set for judgment error, f t (x k ) is the label value predicted by the sub-learner, maxf t (x k )-b k | is the maximum error of the learner on the sample set obtained in the tth iteration, is the relative error of the kth sample. Formula (6) and Formula (7) are the error calculation formulas for the sub-learner classification model and regression model respectively.

[0098] The next step is to set the weight matrix ω of the sub-learner k , during the iteration process, the weight matrix of the sub-learner is specifically:

[0099]

[0100] Where I is the identity matrix.

[0101] The weight matrix values ​​of the training samples are continuously adjusted during the iteration process, specifically:

[0102]

[0103] Among them, G t is a normalization constant. In this embodiment, equations (9) and (10) are weight matrix iteration formulas for the classification model and the regression model.

[0104] The final strong learner is:

[0105]

[0106] Among them, Γ(z t (x k )=m) is the probability that the classification result of the classifier is the mth class, and the argmax function selects the class with the largest probability among all classes as the output of the strong learner. In this embodiment, formula (11) and formula (12) are the sub-learner integration strategies of the classification model and the regression model respectively.

[0107] S104. Evaluate the sewage water body according to the sewage evaluation model, and determine the winery sewage concentration according to the three-dimensional water body model.

[0108] After the strong learner is constructed in this embodiment, it is used as a sewage assessment model to perform water body assessment and obtain the total nitrogen value of the sewage.

[0109] After obtaining the total nitrogen value of the sewage, the volume of the sewage water body is calculated according to the three-dimensional model of the water body. The specific process is as follows:

[0110] Since the shape of the sewage body is irregular, it cannot be directly calculated by the volume calculation method of a regular object. Therefore, this embodiment uses a slice calculation method to calculate the volume of the water body. The specific process is as follows:

[0111] A three-dimensional model of a water body is constructed according to the calculated point cloud data, and the three-dimensional model of the water body is sliced ​​horizontally at equal intervals to obtain a number of slice polygons;

[0112] The α-shape method is used to iteratively find the threshold of each point cloud slice, and the area of ​​the outer contour polygon of the slice is obtained under this threshold. Specifically:

[0113] When using the α-shape method to construct contours, there is a threshold α j , when α≥α j When , the boundary polygon formed can contain all points, so this embodiment uses the α-shape method to iteratively find the threshold α of each point cloud slice j , and the area of ​​the slice outer contour polygon is obtained under this threshold. The specific formula is:

[0114] α j+1 =α j +Δα (13)

[0115] Among them, Δα is the step size and j is the number of iterations.

[0116] The slice area is calculated using the discretized Green's formula as follows:

[0117]

[0118] Among them, S is the slice area of ​​the three-dimensional water model, k is the total number of slice boundary points, (x j ,y j ), is the two-dimensional projection coordinate of point j, (x j +1,y j +1) is the two-dimensional projection coordinate of point j+1.

[0119] After calculating the slice area, the volume V of the three-dimensional model is calculated, specifically:

[0120]

[0121] Among them, S i′ is the slice area of ​​the 3D water model, Δ is the straight-line distance between two adjacent slices, n′ is the number of slices, d i′ is the water body fault coefficient.

[0122] Specifically, in this embodiment, the specific calculation method of the water body fault coefficient is:

[0123] Obtain the plane boundary of the water body slice breakpoint data, and use the plane α-shape edge extraction algorithm to obtain the convex hull boundary of the target fault breakpoint data on the current time slice;

[0124] The polygonal paranoid method with inclination constraints is used to calculate the convex hull expansion polygon and obtain the spatial boundary of the target fault of adjacent time slices;

[0125] Search for the breakpoint data within the convex hull expansion polygon on the adjacent time slices, and determine whether it is an empty set. If so, output the plane shape of the target water body fault. Otherwise, determine whether the breakpoint data belongs to the water body fault. The process of determining whether the breakpoint data belongs to the water body fault is as follows: for the breakpoint data within the convex hull expansion polygon, set the target fault overlap threshold. If the overlap is greater than the threshold, it is determined to be a non-target fault. If it is less than the threshold, it is determined to belong to the target fault.

[0126] If yes, the target water body fault plane shape is output; if no, the adjacent time slice is updated to the current time slice, and the breakpoint data and convex hull boundary are updated;

[0127] The current breakpoint data is merged into the target water body fault, the target water body fault space boundary is merged, the target fault plane morphology is obtained, and the Euclidean distance between the center points of the adjacent target fault spaces is used as the water body fault coefficient.

[0128] Finally, after constructing the volume of the three-dimensional model of the water body, the specific concentration of total nitrogen in the water body was calculated as:

[0129]

[0130] Among them, t is the total nitrogen content data of the water body, and c is the total nitrogen concentration of the water body.

[0131] like Figure 2 As shown, this embodiment also discloses a sewage detection device for three-dimensional construction of assembled pits and point cloud data, including a water body data acquisition module 1, a water body three-dimensional model construction module 2, a sewage assessment model construction module 3 and a sewage pollutant concentration assessment module 4:

[0132] The water body data acquisition module 1 is used to collect the measured data of the winery sewage water body and the laser point cloud data, and perform data processing to obtain the treated sewage water body pollutant data and laser point cloud data;

[0133] The water body three-dimensional model building module 2 is used to build a three-dimensional visualization model of the sewage water body according to the laser point cloud data to obtain a three-dimensional model of the water body;

[0134] The sewage assessment model construction module 3 is used to construct a training set and a test set according to the sewage water pollutant data, select the training set, and construct a sewage assessment model based on the AdaBoost strategy training;

[0135] The wastewater pollutant concentration assessment module 4 is used to assess the winery wastewater water body according to the wastewater assessment model, output the assessment data results, and calculate the winery wastewater pollutant concentration based on the three-dimensional water body model of the assessment data result set.

[0136] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present disclosure is not limited by the order of the actions described, because according to the present disclosure, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily required by the present disclosure.

[0137] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0138] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for three-dimensional construction of assembled pits and sewage detection using point cloud data, characterized in that: The following steps are involved: Collect measured data of winery sewage water bodies and laser point cloud data, and process the data to obtain pollutant data of treated sewage water bodies and laser point cloud data; Constructing a three-dimensional visualization model of the sewage water body according to the laser point cloud data to obtain a three-dimensional model of the water body; Constructing a training set and a test set according to the sewage water pollutant data, selecting a training set, and constructing a sewage assessment model based on AdaBoost strategy training; Evaluate the winery wastewater water body according to the wastewater evaluation model, output the evaluation data results, and calculate the winery wastewater pollutant concentration based on the evaluation data results and the three-dimensional model of the water body; The process of constructing a training set and a test set according to the sewage water pollutant data, selecting a training set, and constructing a sewage assessment model based on AdaBoost strategy training is as follows: Obtain training samples based on the training set and initialize a training sample weight matrix; A sub-learning method is adopted and a sub-learner is obtained by training based on the training set after the weights are initialized; Calculating the training error and weight matrix of the sub-learner; The weight matrix of the training sample is obtained by continuously iterating the training and adjusting according to the training error and the weight matrix; Integrate the weight matrices of sub-learners based on the integration strategy and output a strong learner; The process of evaluating the winery sewage water body according to the sewage evaluation model, outputting the evaluation data results, and calculating the winery sewage pollutant concentration based on the evaluation data results and the water body three-dimensional model is as follows: The total nitrogen value data of the sewage is obtained by evaluating the sewage evaluation model, and the volume of the sewage water body is obtained by calculating the three-dimensional model of the water body; The total nitrogen concentration value in the winery wastewater is calculated based on the total nitrogen value data of the wastewater and the volume of the wastewater; The sewage water volume calculation process is: Slicing the three-dimensional water body model at equal intervals horizontally to obtain a plurality of slice polygons; The α-shape method is used to iteratively find the threshold of each point cloud slice polygon, and the area of ​​the outer contour of the slice polygon is obtained based on the discretized Green's formula under this threshold; Calculating the straight-line distance between the slice polygons, and obtaining the volume of the three-dimensional model according to the straight-line distance and the area calculation; The volume calculation process of the three-dimensional model is as follows: Among them, Δ is the straight-line distance between two adjacent slices, n′ is the number of slices, and d i′ is the water body fault coefficient; The calculation process of the water body fault coefficient is as follows: Obtain the plane boundary of the water body slice breakpoint data, and use the plane α-shape edge extraction algorithm to obtain the convex hull boundary of the target fault breakpoint data on the current time slice; The polygonal paranoid method with inclination constraints is used to calculate the convex hull expansion polygon and obtain the spatial boundary of the target fault of adjacent time slices; Search for the breakpoint data within the convex hull expansion polygon on the adjacent time slices, and determine whether it is an empty set. If so, output the plane shape of the target water body fault. Otherwise, determine whether the breakpoint data belongs to the water body fault. The process of determining whether the breakpoint data belongs to the water body fault is as follows: for the breakpoint data within the convex hull expansion polygon, set the target fault overlap threshold. If the overlap is greater than the threshold, it is determined to be a non-target fault. If it is less than the threshold, it is determined to belong to the target fault. If yes, the target water body fault plane shape is output; if no, the adjacent time slice is updated to the current time slice, and the breakpoint data and convex hull boundary are updated; The current breakpoint data is merged into the target water body fault, the target water body fault space boundary is merged, the target fault plane morphology is obtained, and the Euclidean distance between the center points of the adjacent target fault spaces is used as the water body fault coefficient.

2. The method for three-dimensional construction of prefabricated pits and sewage detection using point cloud data according to claim 1, characterized in that: The process of collecting measured data of winery sewage water body and laser point cloud data, and processing the data to obtain treated sewage water body pollutant data and laser point cloud data is as follows: Several laser point cloud scanners are deployed in the sewage water body to collect laser point cloud data, so as to obtain sewage water body point cloud data; the sewage water body is collected at specific time intervals, and the total nitrogen content in the water body is measured by chromotropic acid spectrophotometry to obtain sewage water body pollutant data.

3. The method for three-dimensional construction of prefabricated pits and sewage detection using point cloud data according to claim 2 is characterized in that: The process of placing a number of laser point cloud scanners in the sewage body to collect laser point cloud data and obtain the sewage body point cloud data is as follows: Using a light intensity regulation circuit and an A / D acquisition module to send an acquisition command to the laser scanner; After receiving the command, the laser scanner sends the collected laser point cloud data to the DSP through the light intensity regulation circuit for buffering; The cached laser point cloud data is sent to the data server to complete the sewage water body point cloud data collection process.

4. The method for three-dimensional construction of prefabricated pits and sewage detection using point cloud data according to claim 1, characterized in that: The construction process of the water body three-dimensional model is as follows: Convert the collected laser point cloud data into a coordinate system; The laser point cloud data after coordinate conversion is imported into 3D Max for processing to construct the model framework, and then three-dimensional modeling is performed to obtain a three-dimensional model of the water body.

5. The method for three-dimensional construction of prefabricated pits and sewage detection using point cloud data according to claim 4 is characterized in that: The process of coordinate system conversion is: Calculating the center of the original point of the laser point cloud data, and calculating the center of the converted laser point cloud data; A covariance matrix is ​​constructed based on the center of the original point of the laser point cloud data and the center of the converted laser point cloud data; A multidimensional matrix is ​​constructed according to the covariance matrix, eigenvalues ​​and eigenvectors are calculated according to the multidimensional matrix, and the eigenvector corresponding to the maximum eigenvalue is selected to form a quaternion; The laser point cloud rotation matrix and the translation matrix are obtained according to the quaternion calculation, and the rotation vector and the translation vector are calculated to determine the laser point cloud data after the coordinate transformation.

6. The method for three-dimensional construction of prefabricated pits and sewage detection using point cloud data according to claim 1, characterized in that: The construction process of the winery is as follows: multiple briefings are conducted during construction; the main body of the cellar pit is excavated by manual excavation or pneumatic pickaxe crushing within a range of 1m around the central column foundation; formwork is supported, concrete rear walls are poured, maintenance and high formwork removal, and wall thickness backfilling are carried out; and functional area construction is carried out.

7. The method for three-dimensional construction of prefabricated pits and sewage detection using point cloud data according to claim 1, characterized in that: The calculation method of the sewage pollutant concentration is: Among them, t is the total nitrogen content data of the water body, c is the total nitrogen concentration of the water body, and V is the volume of the three-dimensional model of the water body.

8. A sewage detection device for three-dimensional construction of assembled pits and point cloud data according to any one of claims 1 to 7, characterized in that: It includes a water body data acquisition module (1), a water body three-dimensional model construction module (2), a sewage assessment model construction module (3) and a sewage pollutant concentration assessment module (4); The water body data acquisition module (1) is used to collect measured data of winery sewage water body and laser point cloud data, and perform data processing to obtain treated sewage water body pollutant data and laser point cloud data; The water body three-dimensional model construction module (2) is used to construct a three-dimensional visualization model of the sewage water body according to the laser point cloud data to obtain a three-dimensional water body model; The sewage assessment model construction module (3) is used to construct a training set and a test set according to the sewage water body pollutant data, select a training set, and construct a sewage assessment model based on AdaBoost strategy training; The wastewater pollutant concentration assessment module (4) is used to assess the winery wastewater water body according to the wastewater assessment model, output the assessment data results, and calculate the winery wastewater pollutant concentration based on the assessment data results and the water body three-dimensional model.

Images