Soil heavy metal content prediction method and soil pretreatment device

Abstract

The invention relates to a soil heavy metal content prediction method and a soil pretreatment device, the soil pretreatment device comprises a cylinder with an input port and an output port respectively formed in two ends, a plurality of cross rods are uniformly distributed in the cylinder in the circumferential direction, a spiral strip is further arranged in the cylinder, and sieve pores are formed by any screw pitch of the spiral strip and the distance between any two adjacent cross rods; and the spiral strip or the cross rod is connected with a driving mechanism. Screening holes are formed through matching of the screw pitches of the spiral strips and the intervals between the multiple transverse rods, so that foundation soil screening operation is achieved, through the spiral structure characteristics of the spiral strips, when the spiral strips or the transverse rods rotate, soil can be conveyed, a filter screen structure and a conveying piece structure are integrated, and meanwhile, the conveying piece structure can be used for conveying soil. Under the action of the spiral strip, one of the spiral strip and the transverse rod rotates, the other one of the spiral strip and the transverse rod is static and gravity, soil is subjected to shearing force of the spiral strip and the transverse rod, and when the soil or sundries are clamped into the screen holes, the soil or the sundries can be crushed or stripped to reduce the condition that the soil or the sundries are retained to block the screen holes, so that the screening efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of soil prediction, and particularly to a method for predicting soil heavy metal content and a soil pretreatment device. Background Art

[0002] Soil heavy metal pollution is an important part of global environmental problems. Among them, vanadium (V), as a trace element, its content in soil not only affects plant growth and ecosystem health, but also is closely related to human health. Therefore, accurately predicting soil heavy metal content is of great significance for environmental risk assessment, agricultural management and pollution control.

[0003] When collecting soil at the sampling point, impurities contained in the soil are often collected together. If these impurities are not processed, it will directly affect the accuracy of soil prediction results. Therefore, the collected soil needs to be screened and filtered before prediction to remove garbage, gravel and other items in the soil, so as to facilitate subsequent prediction by staff.

[0004] However, most of the existing screening devices use a filter screen structure to screen the soil. During the soil screening operation, when the soil flows through the screen layer, some impurities are likely to stay and block the screen holes due to factors such as shape, size or viscosity. This blockage phenomenon directly affects the working efficiency of soil screening. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a method for predicting soil heavy metal content and a soil pretreatment device to solve the problem that most of the existing screening devices use a filter screen structure to screen the soil, and during the soil screening operation, it is easy to stay and block the screen holes, and this blockage phenomenon directly affects the working efficiency of soil screening.

[0006] The present invention is achieved through the following technical solutions:

[0007] A soil pretreatment device includes a cylinder body with an input port and an output port respectively opened at both ends. A plurality of cross bars are evenly distributed along the circumferential direction of the cylinder body. A spiral strip extending in a spiral shape along the axial direction of the cylinder body is further arranged in the cylinder body. The spiral strip is attached to the opposite side of the cross bar. The pitch of the spiral strip and the distance between any two adjacent cross bars form a screen hole. A screening port communicating with its inner cavity is opened on the outer wall of the cylinder body. The screen hole is located between the inner cavity of the cylinder body and the screening port;

[0008] The spiral strip is rotationally matched with the cross bar, and the spiral strip or the cross bar is connected with a driving mechanism.

[0009] Further defined, the cross bar includes an outer rod body and an inner rod body. A gap is formed between the outer rod body and the inner rod body. The opposite sides of the outer rod body and the inner rod body are respectively attached to the opposite sides of the spiral strip.

[0010] Further defined, on the side of the outer rod body and / or the inner rod body close to the spiral strip, a plurality of chutes distributed along the length direction of the cylinder body are provided, sliders are slidably fitted in the chutes, and the sliders are connected to the spiral strip.

[0011] Further defined, a plurality of the chutes are all inclined, and the inclination directions of any two adjacent chutes are opposite.

[0012] Further defined, the pitch of the spiral strip near one end of the output port gradually decreases towards the end of the spiral strip away from the output port.

[0013] Further defined, the screening port includes a plurality of through ports distributed along the axial direction of the cylinder body, a plurality of the through ports are all located on the bottom surface of the cylinder body and any two adjacent through ports are spaced apart, an output funnel is arranged on the through port, and one end of the output funnel away from the tip is connected to the cylinder body.

[0014] Further defined, an elastic scraping piece is arranged on the side surface of the inner rod body, one end of the elastic scraping piece is connected to the inner rod body, and the other end is attached to the surface of the spiral strip.

[0015] A method for predicting soil heavy metal content, including a soil pretreatment device, further includes the following steps:

[0016] S1: Collect the soil to be measured in the target area according to the existing sample point position information in the target area, and pretreat the collected soil through the soil pretreatment device;

[0017] S2: Collect heavy metal data from the pretreated soil, and at the same time obtain the normalized difference vegetation index, saturation vapor pressure deficit, temperature, precipitation, relative humidity, light absorption proportional factor, carbon dioxide concentration, soil sand content, soil clay content, vapor pressure, maximum root depth, drought index, ten-meter high wind speed, soil moisture data through satellite remote sensing spectrum inversion, and perform input standardization processing on the obtained data. Using the mean-standard deviation standardization method, the data is normalized and used as input variables;

[0018] S3: Build a fully connected neural network model, input the input variable data in S2 into the fully connected neural network, extract features and use the training set data to train the model. During the training process, use the validation set to regularly evaluate the model performance. Through the determination coefficient R of the training set and the validation set 2 respectively reach the preset values, confirm that there is no significant overfitting or underfitting phenomenon in the model, and finally obtain a soil heavy metal content prediction model through training;

[0019] S4: Use the features extracted in S3 as independent variables, and calculate through the trained fully connected neural network model, and the output result is the predicted value of the heavy metal content of the soil to be measured.

[0020] Further defined, the network structure of the fully connected neural network model includes four layers:

[0021] The first layer is a fully connected layer with one hundred neurons and the activation function is sigmoid;

[0022] The second layer is a fully connected layer with one hundred neurons and the activation function is sigmoid;

[0023] The third layer is a fully connected layer with one hundred neurons and the activation function is sigmoid;

[0024] The fourth layer is a fully connected layer with one neuron and no activation function.

[0025] Further defined, the coefficient of determination R of the training set and the validation set 2 Formula:

[0026]

[0027] The beneficial effects of the present invention are as follows:

[0028] The sieve holes are formed by the pitch of the spiral bar and the spacing between multiple cross bars, thereby realizing the basic soil screening operation. And due to the spiral structure characteristics of the spiral bar, when the spiral bar or the cross bar rotates, the soil can be conveyed. The filter screen structure and the conveying member structure are integrated. At the same time, when one of the spiral bar and the cross bar rotates and the other is stationary, and under the action of gravity, the soil is subjected to the shear force of the spiral bar and the cross bar. When the soil or sundries are stuck in the sieve holes, they will be crushed or peeled off to reduce the situation of sieve hole blockage caused by their retention, thereby improving the screening efficiency.

[0029] Other advantages, objectives and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a three-dimensional structure schematic diagram of the present invention;

[0031] Figure 2 is a partial structure schematic diagram of the driving structure of the present invention;

[0032] Figure 3 is an internal structure schematic diagram of the cylinder of the present invention;

[0033] Figure 4 is a structure schematic diagram of the cross bar and the spiral bar of the present invention;

[0034] Figure 5 is the top view of the present invention;

[0035] Figure 6 is Figure 6 the sectional view taken along A-A in

[0036] Figure 7 is the structural schematic diagram of the chute and the slider in the second embodiment;

[0037] Figure 8 is the structural schematic diagram of the spiral bar in the third embodiment;

[0038] Figure 9 is the flow chart for constructing the prediction soil V content model of the present invention;

[0039] Figure 10 is the scatter plot of the measured value and prediction of the predicted V content of the present invention.

[0040] In the figure:

[0041] 1, cylinder body; 101, input port; 102, output port; 103, screening port; 2, cross bar; 201, outer rod body; 202, inner rod body; 2021, elastic scraping piece; 3, spiral bar; 4, sieve hole; 5, chute; 501, slider; 6, output funnel; 7, drive motor; 701, first pulley; 702, second pulley; 703, transmission belt; 704, mounting ring. Detailed implementation manners

[0042] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations.

[0043] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0044] It should be noted that: similar reference numerals and letters denote similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings.

[0045] In the above description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "one side" and "the other side" is based on the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the invention is usually placed during use. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, terms such as "first" and "second" are only used for distinguishing descriptions and cannot be construed as indicating or implying relative importance.

[0046] In addition, terms such as "the same" do not mean that the components are required to be absolutely the same, but there may be slight differences. The term "vertical" only means that the positional relationship between components is relatively more vertical compared to "parallel", and does not mean that the structure must be completely vertical, but can be slightly inclined.

[0047] Please refer to Figures 1-10 , the present invention provides a technical solution: a soil pretreatment device, including a cylinder 1 with an input port 101 and an output port 102 respectively opened at both ends. A plurality of cross bars 2 are evenly distributed along the circumference of the cylinder 1. A spiral bar 3 extending in a spiral shape along the axial direction of the cylinder 1 is further arranged in the cylinder 1. One side of the spiral bar 3 is in contact with the cross bar 2. The pitch of any spiral bar 3 and the distance between any two adjacent cross bars 2 form a sieve hole 4. A screening port 103 communicating with its inner cavity is opened on the outer wall of the cylinder 1. The sieve hole 4 is located between the inner cavity of the cylinder 1 and the screening port 103;

[0048] The spiral bar 3 is rotatably matched with the cross bar 2, and the spiral bar 3 or the cross bar 2 is connected with a driving mechanism.

[0049] In this solution, the pitch of any spiral bar 3 and the distance between two adjacent cross bars 2 within the pitch range cooperate to form a sieve hole 4, so as to perform a screening operation on the soil. Since the spiral bar 3 and the cross bar 2 are relatively rotatably matched, and the structure characteristics of the spiral bar 3 are used to guide the soil in the cylinder 1 to make a spiral displacement along the axial direction of the cylinder 1. While forming the sieve hole 4 through the cooperation of the spiral bar 3 and the cross bar 2, the soil can also be conveyed, integrating a filter screen and a conveying member, so as to extend its residence time in the screening area for sufficient screening operation; when the soil is displaced in the cylinder 1, it will be in contact with both the spiral bar 3 and the cross bar 2 at the same time. And due to the rotation of one of the spiral bar 3 and the cross bar 2 and the static state of the other, as well as the action of gravity, the soil is subjected to the shear force of the spiral bar 3 and the cross bar 2. When the soil or debris gets stuck in the sieve hole 4, it will be crushed or peeled off to reduce the blockage of the sieve hole 4 caused by its residence, thereby improving the screening efficiency.

[0050] Among them, the driving mechanism includes a driving motor 7, a first pulley 701 coaxially connected to the output shaft of the driving motor 7, a second pulley 702 rotatably installed at the end of the cylinder body 1, a transmission belt 703 tightly sleeved on the first pulley 701 and the second pulley 702, and a mounting ring 704 also installed at the end of the cylinder body 1. One side of the mounting ring 704 is connected to the second pulley 702, and the opposite side is connected to the spiral strip 3 or the cross bar 2. By starting the driving motor 7, the first pulley 701 and the second pulley 702 are driven to rotate synchronously under the action of the transmission belt 703. Since the mounting ring 704 is coaxially connected to the second pulley 702, the mounting ring 704 drives the spiral strip 3 or the cross bar 2 to rotate in the cylinder body 1 under the transmission of the second pulley 702;

[0051] During specific use, the collected soil is conveyed into the cylinder body 1 through the input port 101. The soil in the cylinder body 1 naturally falls on the side of the spiral strip 3 and the cross bar 2 away from the inner wall of the cylinder body 1. Under the action of gravity, the soil comes into contact with the spiral strip 3 and the cross bar 2. As the driving mechanism drives the spiral strip 3 or the cross bar 2 to rotate, the soil is displaced in the cylinder body 1 under the guidance of the structural characteristics of the spiral strip 3 and is screened in cooperation with the sieve holes 4 during the displacement. The suitable soil is discharged through the sieve holes 4 and the screening port 103, and the sundries are discharged through the output port 102.

[0052] Among them, the size of the sieve holes 4 depends on the pitch of the spiral strip 3 and the spacing of the multiple cross bars 2, which can be adaptively adjusted according to specific use and production to meet various different screening requirements.

[0053] To achieve additional technical effects, the present application can also connect the spiral strip 3 and the cross bar 2 to driving devices respectively to drive both of them to rotate to realize soil screening. Among them, the rotational speeds of the spiral strip 3 and the cross bar 2 can be made different by the driving devices, or the spiral strip 3 and the cross bar 2 rotate in opposite directions through the driving devices. In this application mode, a shearing force can still be formed between the spiral strip 3 and the cross bar 2.

[0054] To achieve better use effects, a vibration generating mechanism (such as an eccentric shaft vibration mechanism) is provided on or integrated with the cylinder body 1. The vibration generating mechanism applies a vibration function to the cylinder body 1. Under the vibration effect, the soil can shake in the cylinder body 1, further improving the screening efficiency;

[0055] The cylinder body 1 is inclined with the input end tilted upward to facilitate the transportation of the soil in the cylinder body 1 under the action of gravity.

[0056] In this embodiment, the cross bar 2 includes an outer rod body 201 and an inner rod body 202. A gap is formed between the outer rod body 201 and the inner rod body 202. The opposite sides of the outer rod body 201 and the inner rod body 202 are respectively attached to the opposite sides of the spiral strip 3.

[0057] Wherein, solid lubricating layers are respectively arranged on one opposite side of the outer rod body 201 and the inner rod body 202, and the solid lubricating layer can adopt polytetrafluoroethylene (PTFE) or molybdenum disulfide (MoS2) coating.

[0058] In this solution, by arranging solid lubricating layers on one opposite side of the outer rod body 201 and the inner rod body 202, when the spiral strip 3 contacts with both of them, the friction coefficient among the three can be reduced through the solid lubricating layer, and the erosion of soil chemicals on the outer rod body 201 and the inner rod body 202 can be reduced, prolonging the service life of the components.

[0059] In this embodiment, a plurality of chutes 5 distributed along the length direction of the cylinder body 1 are arranged on one side of the outer rod body 201 and / or the inner rod body 202 close to the spiral strip 3, and a slider 501 is slidably fitted in the chute 5, and the slider 501 is connected with the spiral strip 3.

[0060] In this solution, when the spiral strip 3 or the cross bar 2 rotates, the slider 501 on the spiral strip 3 rotates in the chute 5, and through the sliding fit of the slider 501 and the chute 5, under the limiting action of the chute 5, the stability of the spiral strip 3 is improved.

[0061] In this embodiment, an elastic scraping piece 2021 is arranged on the side surface of the inner rod body 202, one end of the elastic scraping piece 2021 is connected with the inner rod body 202, and the other end is attached to the surface of the spiral strip 3.

[0062] In this solution, the fixed end of the elastic scraping piece 2021 is connected with the inner rod body 202, and the movable end keeps being in a state of abutting against the surface of the spiral strip 3 under the action of elastic force, and during the continuous rotation of the spiral strip 3 or the cross bar 2, the position where the elastic scraping piece 2021 abuts against the spiral strip 3 continuously changes as the spiral strip 3 or the cross bar 2 rotates, and its movable end scrapes off the soil or sundries on the surface of the spiral strip 3.

[0063] A method for predicting the heavy metal content in soil, including a soil pretreatment device, and further comprising the following steps:

[0064] S1: Collect the soil to be measured in the target area according to the existing sample point position information in the target area, and pretreat the collected soil through the soil pretreatment device;

[0065] S2: Collect data on heavy metals (taking vanadium (V) as an example in this application) from the pre-treated soil. At the same time, obtain the normalized difference vegetation index, vapor pressure deficit, temperature, precipitation, relative humidity, light absorption ratio factor, carbon dioxide concentration, soil sand content, soil clay content, vapor pressure, maximum root depth, drought index, ten-meter high wind speed, and soil moisture data through satellite remote sensing spectral inversion. Input standardization processing is performed on the obtained data. Using the mean-standard deviation standardization method, the data is normalized and used as input variables;

[0066] S3: Build a fully connected neural network model, and input the input variable data in S2 into the fully connected neural network to extract features and use the training set data to train the model. During the training process, use the validation set to regularly evaluate the model performance. Through the determination coefficient R of the training set and the validation set 2 reach the preset values respectively, confirm that there is no significant overfitting or underfitting phenomenon in the model, and finally obtain the soil heavy metal V content prediction model through training;

[0067] S4: Take the features extracted in S3 as independent variables and calculate through the trained fully connected neural network model. The output result is the predicted value of the heavy metal V content in the soil to be measured.

[0068] Among them, the specific training process of the fully connected neural network model is as follows: Use the training set data (630 samples) to train the model. The number of training epochs is 1000, the optimizer is Adam, the learning rate is 0.01, the regularization type is L2, the regularization rate is 0.001, and there is no random dropout and cyclic random dropout; Calculate the loss function according to the output value and the true value, and use the backpropagation algorithm to update the model parameters until the model performance converges;

[0069] During the training process, use the validation set (270 samples) to regularly evaluate the model performance. The determination coefficient R of the training set 2 reaches 0.984, and the determination coefficient R of the validation set 2 reaches 0.565, and confirm that there is no significant overfitting or underfitting phenomenon in the model.

[0070] In this solution, impurities are efficiently removed through the pre-treatment device to ensure the purity of the soil samples, fundamentally eliminating the interference of debris on the detection of heavy metal V. Combining multi-source environmental parameters (such as climate, soil physical and chemical properties, etc.), a more comprehensive prediction factor system is constructed to improve the accuracy of predicting soil heavy metal content.

[0071] In S2, the dimension difference is eliminated through input data standardization processing to improve the model convergence speed; the fully connected neural network deeply extracts the non-linear relationship between features and adapts to the heavy metal migration law under complex environments.

[0072] That is, through this method for predicting soil heavy metal content, multi-source dynamic data such as the normalized difference vegetation index, vapor pressure deficit, and drought index are integrated, as well as variables related to heavy metal V in the soil, fully capturing the heavy metal V content and soil factors, and improving the comprehensiveness and accuracy of the prediction.

[0073] By using the non-linear feature extraction ability of the fully connected neural network and combining input standardization and L2 regularization techniques, the problems of insufficient comprehensive utilization of multi-source data and model overfitting in traditional methods are effectively alleviated.

[0074] The results of model training and validation show that the R value of the training set 2 is 0.984, and the R value of the validation set 2 is 0.565, with a relatively high prediction accuracy, and it can achieve accurate prediction of soil heavy metal content.

[0075] In this embodiment, the network structure of the fully connected neural network model includes four layers:

[0076] The first layer is a fully connected layer with one hundred neurons and the activation function is sigmoid;

[0077] The second layer is a fully connected layer with one hundred neurons and the activation function is sigmoid;

[0078] The third layer is a fully connected layer with one hundred neurons and the activation function is sigmoid;

[0079] The fourth layer is a fully connected layer with one neuron and no activation function.

[0080] In this solution, high-dimensional features are fully captured through three hidden layers (one hundred neurons in each layer), and the sigmoid function provides non-linear fitting ability, while the output layer without an activation function ensures that the predicted value is continuous and unbounded, conforming to the characteristics of the heavy metal content range.

[0081] In this embodiment, the coefficient of determination R of the training set and the validation set 2 Formula:

[0082]

[0083] Among them, y i represents the true value of sample i, represents the predicted value of sample i, represents the average value of sample i.

[0084] Embodiment 2: The difference from Embodiment 1 is that multiple said sliding grooves 5 are all inclined, and the inclination directions of any two adjacent said sliding grooves 5 are opposite.

[0085] Among them, the spiral strip 3 is made of an elastic material and can stretch or contract along the axial direction of the cylinder 1.

[0086] In this embodiment, the crossbar 2 is rotated by a driving mechanism, while the spiral strip 3 does not rotate, and a plurality of sliders 501 are provided on the surface of the spiral strip 3. The plurality of sliders 501 are distributed axially on the spiral strip 3 and also along the spiral direction of the spiral strip 3. The plurality of sliders 501 can be rotated into the chute 5 in sequence by rotating the crossbar 2 and slidingly engaged with the chute 5. The plurality of sliders 501 can also be rotated out of the chute 5 by rotating the crossbar 2.

[0087] During specific use, the driving mechanism drives the crossbar 2 to rotate. As the crossbar 2 rotates, the chutes 5 on the inner rod body 202 and the outer rod body 201 also rotate accordingly. The rotation of the chutes 5 causes the sliders 501 on the spiral strip 3 to sequentially enter the chutes 5 and slide with them, causing the two sliders 501 to slide along the inclined directions of the two chutes 5 respectively. Since the inclined directions of the two adjacent chutes 5 are opposite, the spiral strip 3 will present the following state:

[0088] State 1: When two adjacent chutes 5 are inclined toward each other, two adjacent sliders 501 on the spiral strip 3 enter the two chutes 5 respectively. As the two chutes 5 continue to rotate, the two sliders 501 slide toward each other, causing the area where the spiral strip 3 connects with the two sliders 501 to shrink and deform, and the sieve holes 4 matched therewith will shrink.

[0089] State 2: When two adjacent chutes 5 are tilted away from each other, two adjacent sliders 501 on the spiral strip 3 enter the two chutes 5 respectively. As the two chutes 5 continue to rotate, the two sliders 501 slide away from each other, causing the area where the spiral strip 3 connects with the two sliders 501 to stretch and deform, and the sieve holes 4 matched with them to increase in size.

[0090] State three: As the chute 5 continues to rotate, when the two sliders 501 rotate to disengage from the chute 5, the deformed area of the spiral strip 3 is reset under the elastic action of the spiral strip 3 itself, and the sieve hole 4 matched with it is restored to its original state. As the deformed area of the spiral strip 3 is reset, the remaining sliders 501 can enter the chute 5 by rotating, thereby realizing intermittent deformation and restoration of the spiral strip 3.

[0091] That is, by rotating the crossbar 2, with the cooperation of the chute 5 and the slider 501, the spiral bar 3 is deformed, and the aperture of the sieve hole 4 also changes accordingly. The soil is screened by the dynamically changing sieve hole 4, which significantly improves the prevention of retention and clogging. The reduction of the sieve hole 4 can produce a squeezing effect on the soil, and can crush part of the soil.

[0092] Embodiment 3: The difference from Embodiment 1 is that the pitch of the spiral bar 3 near one end of the outlet 102 tapers towards the end of the spiral bar 3 away from the outlet 102.

[0093] In this solution, the spiral bar 3 is rotated by a driving mechanism, while the cross bar 2 does not rotate. The spiral bar 3 is used to convey the soil. The soil entering the cylinder 1 through the input port 101 first contacts the end of the spiral bar 3 with a smaller pitch, and is conveyed towards the outlet 102 along with the spiral bar 3. The soil gradually contacts the end of the spiral bar 3 with a larger pitch. During this process, soils of different sizes can be sieved in sequence to achieve classification sieving through the tapered pitch.

[0094] In this embodiment, the screening port 103 includes a plurality of through ports distributed along the axial direction of the cylinder 1. All of the plurality of through ports are located on the bottom surface of the cylinder 1 and any two adjacent through ports are spaced apart. An output funnel 6 is provided on the through port, and one end of the output funnel 6 away from the tip is connected to the cylinder 1.

[0095] The screening port 103 is separated into a plurality of through ports, and the plurality of through holes correspond to different pitch segments of the spiral bar 3, so as to separate different pitch segments of the spiral bar 3 into a plurality of classification screening areas and output them respectively through the plurality of through ports, facilitating subsequent targeted processing.

[0096] To achieve a better use effect, a plurality of collection containers can be correspondingly arranged below the plurality of through ports.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the purpose and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A soil pretreatment device, characterized in that: It includes a cylinder body (1) with an input port (101) and an output port (102) respectively opened at both ends. A plurality of cross bars (2) are evenly distributed along the circumferential direction of the cylinder body (1). A spiral strip (3) extending spirally along the axial direction of the cylinder body (1) is further arranged in the cylinder body (1). One side of the spiral strip (3) is attached to the opposite side of the cross bar (2). The pitch of any part of the spiral strip (3) and the distance between any two adjacent cross bars (2) form a sieve hole (4). A screening port (103) communicating with its inner cavity is opened on the outer wall of the cylinder body (1). The sieve hole (4) is located between the inner cavity of the cylinder body (1) and the screening port (103). The spiral strip (3) is rotationally matched with the cross bar (2), and the spiral strip (3) or the cross bar (2) is connected with a driving mechanism.

2. The soil pretreatment device according to claim 1, characterized in that: The cross bar (2) includes an outer rod body (201) and an inner rod body (202). A gap is formed between the outer rod body (201) and the inner rod body (202). The opposite sides of the outer rod body (201) and the inner rod body (202) are respectively attached to the opposite sides of the spiral strip (3).

3. The soil pretreatment device according to claim 2, wherein: On the side of the outer rod body (201) and / or the inner rod body (202) close to the spiral strip (3), a plurality of sliding grooves (5) distributed along the length direction of the cylinder body (1) are arranged. A slider (501) is slidably matched in the sliding groove (5), and the slider (501) is connected with the spiral strip (3).

4. The soil pretreatment device according to claim 3, characterized in that: A plurality of the sliding grooves (5) are all inclined, and the inclination directions of any two adjacent sliding grooves (5) are opposite.

5. The soil pretreatment device according to claim 1, characterized in that: The pitch of the end of the spiral strip (3) close to the output port (102) gradually shrinks towards the end of the spiral strip (3) far from the output port (102).

6. The soil pretreatment device according to claim 5, characterized in that: The screening port (103) includes a plurality of through ports distributed along the axial direction of the cylinder body (1). A plurality of the through ports are all located on the bottom surface of the cylinder body (1) and any two adjacent through ports are arranged at intervals. An output funnel (6) is arranged on the through port, and one end of the output funnel (6) far from the tip is connected with the cylinder body (1).

7. The soil pretreatment device according to claim 2, characterized in that: An elastic scraping piece (2021) is arranged on the side surface of the inner rod body (202). One end of the elastic scraping piece (2021) is connected with the inner rod body (202), and the other end is attached to the surface of the spiral strip (3).

8. A method for predicting the heavy metal content in soil, comprising the soil pretreatment device according to any one of claims 1-7, characterized in that: It also includes the following steps: S1: Collect the soil to be measured in the target area according to the existing sample point position information in the target area, and preprocess the collected soil through a soil pretreatment device; S2: Collect heavy metal data of the preprocessed soil. At the same time, obtain the normalized difference vegetation index, saturation vapor pressure deficit, temperature, precipitation, relative humidity, light absorption ratio factor, carbon dioxide concentration, soil sand content, soil clay content, vapor pressure, maximum root depth, drought index, ten-meter high wind speed, soil moisture data through satellite remote sensing spectral inversion, and perform input standardization processing on the obtained data. Using the mean-standard deviation standardization method, normalize the data and use it as the input variable; S3: Build a fully connected neural network model, input the input variable data in S2 into the fully connected neural network, extract features and use the training set data to train the model. During the training process, use the validation set to regularly evaluate the model performance. When the coefficient of determination R of the training set and the validation set 2 respectively reach the preset values, confirm that there is no significant overfitting or underfitting phenomenon in the model, and finally obtain the soil heavy metal content prediction model through training; S4: Use the features extracted in S3 as independent variables, and perform calculations through the trained fully connected neural network model. The output result is the predicted value of the heavy metal content of the soil to be measured.

9. The soil heavy metal content prediction method according to claim 8, characterized in that: The network structure of the fully connected neural network model includes four layers: The first layer is a fully connected layer with one hundred neurons and the activation function is sigmoid; The second layer is a fully connected layer with one hundred neurons and the activation function is sigmoid; The third layer is a fully connected layer with one hundred neurons and the activation function is sigmoid; The fourth layer is a fully connected layer with one neuron and no activation function.

10. The soil heavy metal content prediction method according to claim 8, characterized in that: The coefficient of determination R of the training set and the validation set 2 Formula:

Images