A method for assessing exposure of hazardous waste resource products

Abstract

The present invention discloses a method for assessing the exposure of hazardous waste resource products, including collecting production processes, application scenarios and physical and chemical test data of the whole process of hazardous waste resource products; inputting the physical and chemical test data of the whole process into a toxicity prediction model to obtain toxicity data, classifying the production process and the application scenarios to obtain impact data; obtaining an impact factor according to the correlation of the impact data to the exposure, and establishing a risk index according to the toxicity data and the exposure; constructing an exposure assessment model based on the toxicity data, the impact factor and the risk index, optimizing the parameters of the exposure assessment model, inputting the data to be assessed into the exposure assessment model, and outputting an assessment result. This method can not only improve the accuracy of the exposure assessment of hazardous waste resource products, but also help to better manage and control hazardous waste resource products, achieve optimal utilization of resources, and reduce negative impacts on the environment.

Description

Technical Field

[0001] The present invention relates to the technical field of environmental protection and waste resource utilization, and in particular to a method for assessing the exposure of hazardous waste resource utilization products. Background Art

[0002] With the rapid development of industry, the output of hazardous waste is increasing, which poses a serious threat to the environment and human health. The resource utilization of hazardous waste has become an important solution, but in this process, the exposure assessment of resource products is crucial. However, the existing exposure assessment methods have many shortcomings, such as low accuracy and poor interpretability, which make it difficult to effectively meet actual needs. Therefore, there is an urgent need for a more accurate and well-interpretable exposure assessment method for hazardous waste resource products to ensure the safe use of resource products, reduce their potential risks to the environment and human body, and promote the rational resource utilization of hazardous waste. Summary of the invention

[0003] The purpose of the present invention is to provide a method for assessing the exposure of hazardous waste resource products.

[0004] To achieve the above object, the present invention is implemented according to the following technical solutions:

[0005] The present invention comprises the following steps:

[0006] Collect production process, application scenarios and physical and chemical testing data of hazardous waste resource products;

[0007] Inputting the physical and chemical test data of the whole process into a toxicity prediction model to obtain toxicity data, and classifying the production process and the application scenario to obtain impact data;

[0008] Obtaining an impact factor according to the correlation between the impact data and the exposure, and establishing a risk index according to the toxicity data and the exposure;

[0009] An exposure assessment model is constructed based on the toxicity data, the impact factor and the risk index, the parameters of the exposure assessment model are optimized, the data to be assessed are input into the exposure assessment model, and the assessment result is output.

[0010] Furthermore, the method of collecting the production process, application scenarios and physical and chemical testing data of the whole process of hazardous waste resource products includes:

[0011] The physical and chemical testing data of the whole process include the physical and chemical testing data of hazardous wastes, hazardous waste resource products, and by-products in the application scenarios of hazardous waste resource products. The physical and chemical testing data include the content of heavy metals and organic pollutants.

[0012] Furthermore, the method of inputting the whole process physical and chemical test data into the toxicity prediction model to obtain toxicity data includes:

[0013] The pre-trained toxicity prediction model obtains the median lethal dose, carcinogenic risk, and teratogenicity of pollutants through the pollutant dictionary and calculates the toxicity coefficient:

[0014]

[0015] Where T is the toxicity coefficient, LD50 is the median lethal dose, CP is the carcinogenic risk, J is the teratogenicity, and w 1 、w 2 and w 3 They are the weights of LD50, CP and J respectively, and the toxicity data include the toxicity coefficient and concentration of the pollutants.

[0016] Furthermore, the method of classifying the production process and the application scenario to obtain the impact data includes:

[0017] One-hot encoding is used to convert production processes and application scenarios into numerical vectors, and cluster centers are randomly selected to cluster production processes and application scenarios. The initial update formula of cluster centers is:

[0018]

[0019] where v i is the cluster center of cluster i, ω l and ω b are the weights of the lower approximation set and the boundary region, ω ll and ω lb is the weight of the lower approximation set and the boundary area after adding the imbalance adjustment parameter, x j is the jth data point, h ij is the membership of data point j to cluster i, m is the fuzzy coefficient, C i is cluster i, C i is the lower approximation set of cluster i, is the boundary area, is the upper approximation set of cluster i, is an empty set,

[0020] The membership formula of the newly added data point to cluster i is:

[0021]

[0022] in, To add new data points The number of data points belonging to cluster i in the neighborhood data of yes The number of neighborhood data points within the neighborhood radius ζ, for and v i The spacing, Represents data points The local density of k is the number of clusters,

[0023] The update formula of cluster center after adding new data points:

[0024]

[0025] in, for The number of data points in is the number of new data points.

[0026] Furthermore, the method for obtaining the impact factor according to the correlation between the impact data and the exposure degree includes:

[0027] The exposure pathway is obtained based on the impact data, and the formula for calculating the exposure degree is:

[0028]

[0029] Where κ is the exposure, is the number of exposure pathways, is the amount of pollutants, For the pollutant in the first exposure pathway The concentration, F l is the exposure frequency of the first exposure route, D l is the exposure duration of the first exposure route, is the exposure efficiency of the first exposure route, B l is the bioavailability of the first exposure route, AT l is the average exposure time of the first exposure route, Pollutants for the first exposure route Environmental concentration,

[0030] Using exposure as the target value, toxicity data and impact data as input vectors, the SVR model is trained.

[0031] The Bayesian optimization algorithm is used to optimize the hyperparameters of the SVR model. The objective function of the SVR model is:

[0032]

[0033] Where f is the objective function, v is the weight vector of the SVR model, b is the bias term, M is the number of samples, and a i is the i-th input vector, e iis the i-th target value, ∈ is the threshold of insensitive loss, ∈∈(0,0.01), λ is the regularization coefficient, ρ is the weight of the regularization term, ρ∈[0,1], ‖·‖ represents the Euclidean distance,

[0034] The Gaussian process used in Bayesian optimization is:

[0035]

[0036] in is the objective function at the hyperparameter point The value at , gp is the probability distribution of the Gaussian process, is the kernel function, and is a hyperparameter point in the hyperparameter search space, the Gaussian process mean is 0, and the kernel function is a Gaussian kernel function.

[0037] Collection function:

[0038]

[0039] in is the predicted mean, is the prediction standard deviation, θ is the hyperparameter search space of the SVR model, β is the control coefficient of the control acquisition strategy, and the update rule of the prediction distribution is:

[0040]

[0041] in According to D 1:t and The objective function is obtained in The predicted distribution at The mean and variance are and Gaussian distribution, is the objective function in the tth iteration The prediction variance at 1:t is the set of hyperparameter points for the 1st to tth iterations, t is the current number of iterations, and f 1:t D 1:t The corresponding objective function Given a new hyperparameter point at the tth iteration The predicted mean at , is the expected value predicted using the Gaussian process, and K is the t hyperparameter points from the 1st to the tth iteration. The kernel matrix, K -1 is the inverse matrix of K, k is the kernel vector of the hyperparameter point at the t+1th iteration;

[0042] Iterate the optimization until the objective function value is less than 0.05, calculate the SHAP value of the input vector for the target value as the correlation, and take the correlation of the influencing data as the influencing factor of the corresponding hazardous waste resource products.

[0043] Furthermore, a method for establishing a risk index based on the toxicity data and the exposure degree comprises:

[0044] The formula for calculating the risk index is:

[0045]

[0046] Among them, C 风险 is the risk index, is the amount of pollutants, For the The toxicity coefficient of the pollutant, κ is the exposure, For the The exposure concentration of the pollutant is the total hazard quotient, The non-carcinogenic hazards of these pollutants in all exposure pathways, including corrosiveness, neurotoxicity and ecotoxicity, For the The carcinogenic risk of a given exposure route is given by e, where e is the base of the natural logarithm.

[0047] Furthermore, a method for constructing an exposure assessment model based on the toxicity data, the impact factor and the risk index comprises:

[0048] The toxicity data, influencing factors and risk index are divided into training set, test set and validation set. The training set is used to learn the prediction pattern of toxicity data and influencing factors for risk index based on the Xgboost algorithm, and an exposure assessment model is established. The test set is used to optimize the hyperparameters of the exposure assessment model, and the accuracy of the exposure assessment model is verified by the validation set.

[0049] Furthermore, the method for optimizing the exposure assessment model comprises:

[0050] The hyperparameters of the exposure assessment model are used as particles, and the mean square error of the exposure assessment model is used as the fitness value of the particles. The particles are ranked from small to large according to the fitness value, and the particle mass and inertial mass are calculated according to the ranking:

[0051]

[0052] Where m i (t) and m j (t) are the masses of the i-th particle and the j-th particle in the t-th iteration, respectively. i (t) is the rank of particle i at iteration t, N is the total number of particles,

[0053] The formula for calculating the gravitational force between particles is:

[0054]

[0055] in is the gravitational force on the ith particle in the tth iteration, ω j is the weight of the jth particle, ranging from [0,1], is the gravitational force exerted on the i-th particle in the d-th dimension by the j-th particle in the t-th iteration, G(t) is the gravitational constant in the t-th iteration, e is the base of the natural logarithm, α(t) is the attenuation coefficient of the tth iteration, T is the maximum number of iterations, M i (t) and M j (t) are the inertial masses of the ith particle and the jth particle in the tth iteration, R ij (t) is the Euclidean distance between the ith particle and the jth particle in the tth iteration, ε is the allowable error, and are the positions of the ith particle and the jth particle in the dth dimension in the tth iteration,

[0056] Dynamically adjust the attenuation coefficient of the gravitational constant and the initial gravitational constant according to the number of iterations:

[0057]

[0058]

[0059] Among them, α 0 is the initial attenuation coefficient, β is the adjustment parameter, G 0 (t) is the initial gravitational constant of the tth iteration, G max and G min are the maximum and minimum values ​​of the initial gravitational constant,

[0060] Particle velocity update formula and position update formula:

[0061]

[0062]

[0063] in and are the velocities of the i-th particle in d dimension at the t-th iteration and the t+1-th iteration, respectively. is a random number between [0,1], is the acceleration of the ith particle in the d dimension at the tth iteration, f i(t) is the fitness value of particle i in the tth iteration, k v is the speed adjustment factor, f best (t) and f worst (t) are the minimum and maximum fitness values ​​in the tth iteration, is the position of the optimal solution in the dth dimension in the tth iteration. The optimal solution is the particle with the smallest fitness value. is the position of the i-th particle in the d-dimensional space at the t+1-th iteration. n(t) is a random noise term, which is a Gaussian noise with zero mean and time-varying variance. The optimal solution is output after the preset maximum number of iterations is reached.

[0064] In a second aspect, an embodiment of the present application further provides an electronic device, including:

[0065] A processor; and a memory arranged to store computer executable instructions, which when executed cause the processor to perform the method steps described in the first aspect.

[0066] In a third aspect, an embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs. When the one or more programs are executed by an electronic device including multiple applications, the electronic device executes the method steps described in the first aspect.

[0067] The beneficial effects of the present invention are:

[0068] The present invention is a method for assessing the exposure of hazardous waste resource products. Compared with the prior art, the present invention has the following technical effects:

[0069] The present invention can more accurately assess the exposure of hazardous waste resource products by comprehensively considering toxicity data, production processes and application scenarios, and provide a more reliable basis for the resource utilization of hazardous waste. This method can be directly applied to the exposure assessment system of hazardous waste resource products, providing scientific support for relevant decision-making, and helping to improve the safety and rationality of hazardous waste resource utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 A flowchart of the steps of a method for assessing exposure of hazardous waste resource products of the present invention;

[0071] Figure 2 It is a schematic diagram of the structure of an electronic device in an embodiment of this specification. DETAILED DESCRIPTION

[0072] The present invention is further described below by means of specific embodiments. The illustrative embodiments and descriptions of the present invention are used to explain the present invention but are not intended to limit the present invention.

[0073] The present invention provides a method for assessing exposure of hazardous waste resource products, comprising the following steps:

[0074] like Figure 1 As shown, in this embodiment, the following steps are included:

[0075] Collect production process, application scenarios and physical and chemical testing data of hazardous waste resource products;

[0076] In the actual evaluation, the hazardous waste resource products selected were water treatment agents prepared from waste acid. The data collected for three groups of water treatment agents included:

[0077]

[0078]

[0079] The order of the physical and chemical test data in the table is the concentration of Pb, Cr, Cd, Hg and As, and the unit of concentration is mol / L;

[0080] Inputting the physical and chemical test data of the whole process into a toxicity prediction model to obtain toxicity data, and classifying the production process and the application scenario to obtain impact data;

[0081] In the actual evaluation, the toxicity coefficients are: Pb, 0.7, Cr, 0.6, Cd, 0.9, Hg, 1.2 and As, 0.8, and the impact data are: Production process: The production process of water treatment agents 1 and 2 is reactive, and the production process of water treatment agent 3 is synthetic; Application scenario: The application scenario of water treatment agents 1 and 3 is domestic sewage treatment, and the application scenario of water treatment agent 2 is industrial wastewater treatment;

[0082] Obtaining an impact factor according to the correlation between the impact data and the exposure, and establishing a risk index according to the toxicity data and the exposure;

[0083] In the actual evaluation, the impact factors of water treatment agents 1, 2, and 3 were -0.62, -0.53, and -0.58, respectively, and the risk indexes of water treatment agents 1, 2, and 3 were 0.42, 0.52, and 0.45, respectively;

[0084] An exposure assessment model is constructed based on the toxicity data, the impact factor and the risk index, the parameters of the exposure assessment model are optimized, the data to be assessed are input into the exposure assessment model, and the assessment result is output.

[0085] In this embodiment, the method for collecting production process, application scenarios and physical and chemical testing data of the whole process of hazardous waste resource products includes:

[0086] The physical and chemical testing data of the whole process include the physical and chemical testing data of hazardous wastes, hazardous waste resource products, and by-products in the application scenarios of hazardous waste resource products. The physical and chemical testing data include the content of heavy metals and organic pollutants.

[0087] In this embodiment, the method of inputting the whole process physical and chemical test data into the toxicity prediction model to obtain toxicity data includes:

[0088] The pre-trained toxicity prediction model obtains the median lethal dose, carcinogenic risk, and teratogenicity of pollutants through the pollutant dictionary and calculates the toxicity coefficient:

[0089]

[0090] Where T is the toxicity coefficient, LD50 is the median lethal dose, CP is the carcinogenic risk, J is the teratogenicity, and w 1 、w 2 and w 3 They are the weights of LD50, CP and J respectively, and the toxicity data include the toxicity coefficient and concentration of the pollutants.

[0091] In this embodiment, the method of classifying the production process and the application scenario to obtain the impact data includes:

[0092] One-hot encoding is used to convert production processes and application scenarios into numerical vectors, and cluster centers are randomly selected to cluster production processes and application scenarios. The initial update formula of cluster centers is:

[0093]

[0094] where v i is the cluster center of cluster i, ω l and ω b are the weights of the lower approximation set and the boundary region, ω ll and ω lb is the weight of the lower approximation set and the boundary area after adding the imbalance adjustment parameter, x j is the jth data point, h ij is the membership of data point j to cluster i, m is the fuzzy coefficient, C i is cluster i, C i is the lower approximation set of cluster i, is the boundary area, is the upper approximation set of cluster i, is an empty set,

[0095] The membership formula of the newly added data point to cluster i is:

[0096]

[0097] in, To add new data points The number of data points belonging to cluster i in the neighborhood data of yes The number of neighborhood data points within the neighborhood radius ζ, for and v i The spacing, Represents data points The local density of k is the number of clusters,

[0098] The update formula of cluster center after adding new data points:

[0099]

[0100] in, for The number of data points in is the number of new data points.

[0101] In this embodiment, the method of obtaining the impact factor according to the correlation between the impact data and the exposure degree includes:

[0102] The exposure pathway is obtained based on the impact data, and the formula for calculating the exposure degree is:

[0103]

[0104] Where κ is the exposure, is the number of exposure pathways, is the amount of pollutants, For the pollutant in the first exposure pathway The concentration, F l is the exposure frequency of the first exposure route, D l is the exposure duration of the first exposure route, is the exposure efficiency of the first exposure route, B l is the bioavailability of the first exposure route, AT l is the average exposure time of the first exposure route, Pollutants for the first exposure route Environmental concentration,

[0105] Using exposure as the target value, toxicity data and impact data as input vectors, the SVR model is trained.

[0106] The Bayesian optimization algorithm is used to optimize the hyperparameters of the SVR model. The objective function of the SVR model is:

[0107]

[0108] Where f is the objective function, v is the weight vector of the SVR model, b is the bias term, M is the number of samples, and a i is the i-th input vector, e i is the i-th target value, ∈ is the threshold of insensitive loss, ∈∈(0,0.01), λ is the regularization coefficient, ρ is the weight of the regularization term, ρ∈[0,1], ‖·‖ represents the Euclidean distance,

[0109] The Gaussian process used in Bayesian optimization is:

[0110]

[0111] in is the objective function at the hyperparameter point The value at , gp is the probability distribution of the Gaussian process, is the kernel function, and is a hyperparameter point in the hyperparameter search space, the Gaussian process mean is 0, and the kernel function is a Gaussian kernel function.

[0112] Collection function:

[0113]

[0114] in is the predicted mean, is the prediction standard deviation, θ is the hyperparameter search space of the SVR model, β is the control coefficient of the control acquisition strategy, and the update rule of the prediction distribution is:

[0115]

[0116] in According to D 1:t and The objective function is obtained in The predicted distribution at The mean and variance are and Gaussian distribution, is the objective function in the tth iteration The prediction variance at 1:t is the set of hyperparameter points for the 1st to tth iterations, t is the current number of iterations, and f 1:t D 1:t The corresponding objective function Given a new hyperparameter point at the tth iteration The predicted mean at , is the expected value predicted using the Gaussian process, and K is the t hyperparameter points from the 1st to the tth iteration. The kernel matrix, K-1 is the inverse matrix of K, k is the kernel vector of the hyperparameter point at the t+1th iteration;

[0117] Iterate the optimization until the objective function value is less than 0.05, calculate the SHAP value of the input vector for the target value as the correlation, and take the correlation of the influencing data as the influencing factor of the corresponding hazardous waste resource products.

[0118] In this embodiment, the method of establishing a risk index based on the toxicity data and the exposure degree includes:

[0119] The formula for calculating the risk index is:

[0120]

[0121] Among them, C 风险 is the risk index, is the amount of pollutants, For the The toxicity coefficient of the pollutant, κ is the exposure, For the The exposure concentration of the pollutant is the total hazard quotient, The non-carcinogenic hazards of these pollutants in all exposure pathways, including corrosiveness, neurotoxicity and ecotoxicity, For the The carcinogenic risk of a given exposure route is given by e, where e is the base of the natural logarithm.

[0122] In this embodiment, the method for constructing an exposure assessment model based on the toxicity data, the impact factor and the risk index includes:

[0123] The toxicity data, influencing factors and risk index are divided into training set, test set and validation set. The training set is used to learn the prediction pattern of toxicity data and influencing factors for risk index based on the Xgboost algorithm, and an exposure assessment model is established. The test set is used to optimize the hyperparameters of the exposure assessment model, and the accuracy of the exposure assessment model is verified by the validation set.

[0124] In this embodiment, the method for optimizing the exposure assessment model includes:

[0125] The hyperparameters of the exposure assessment model are used as particles, and the mean square error of the exposure assessment model is used as the fitness value of the particles. The particles are ranked from small to large according to the fitness value, and the particle mass and inertial mass are calculated according to the ranking:

[0126]

[0127] Where m i (t) and m j(t) are the masses of the i-th particle and the j-th particle in the t-th iteration, respectively. i (t) is the rank of particle i at iteration t, N is the total number of particles,

[0128] The formula for calculating the gravitational force between particles is:

[0129]

[0130] in is the gravitational force on the ith particle in the tth iteration, ω j is the weight of the jth particle, ranging from [0,1], is the gravitational force exerted on the i-th particle in the d-th dimension by the j-th particle in the t-th iteration, G(t) is the gravitational constant of the t-th iteration, G(t) = G 0 (t)·e -α(t)·t / T , e is the base of the natural logarithm, α(t) is the attenuation coefficient of the tth iteration, T is the maximum number of iterations, M i (t) and M j (t) are the inertial masses of the ith particle and the jth particle in the tth iteration, R ij (t) is the Euclidean distance between the ith particle and the jth particle in the tth iteration, ε is the allowable error, and are the positions of the ith particle and the jth particle in the dth dimension in the tth iteration,

[0131] Dynamically adjust the attenuation coefficient of the gravitational constant and the initial gravitational constant according to the number of iterations:

[0132]

[0133] Among them, α 0 is the initial attenuation coefficient, β is the adjustment parameter, G 0 (t) is the initial gravitational constant of the tth iteration, G max and G min are the maximum and minimum values ​​of the initial gravitational constant,

[0134] Particle velocity update formula and position update formula:

[0135]

[0136] in and are the velocities of the i-th particle in d dimension at the t-th iteration and the t+1-th iteration, respectively. is a random number between [0,1], is the acceleration of the ith particle in the d dimension at the tth iteration, f i (t) is the fitness value of particle i in the tth iteration, k v is the speed adjustment factor, f best (t) and f worst (t) are the minimum and maximum fitness values ​​in the tth iteration, is the position of the optimal solution in the dth dimension in the tth iteration. The optimal solution is the particle with the smallest fitness value. is the position of the i-th particle in the d-dimensional space at the t+1-th iteration. n(t) is a random noise term, which is a Gaussian noise with zero mean and time-varying variance. The optimal solution is output after the preset maximum number of iterations is reached.

[0137] Figure 2 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. Figure 2 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0138] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0139] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0140] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a hazardous waste resource product exposure assessment device at the logical level. The processor executes the program stored in the memory and is specifically used to execute any of the aforementioned hazardous waste resource product exposure assessment methods.

[0141] The above application Figure 1 The method for assessing the exposure of hazardous waste resource products disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a storage medium mature in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps of the above method in combination with its hardware.

[0142] The electronic device may also perform Figure 1 A method for assessing the exposure of hazardous waste resource products was developed and implemented Figure 1 The functions of the illustrated embodiment will not be described in detail in the embodiments of the present application.

[0143] An embodiment of the present application also proposes a computer-readable storage medium, which stores one or more programs, and the one or more programs include instructions. When the instructions are executed by an electronic device including multiple application programs, any one of the aforementioned hazardous waste resource product exposure assessment methods is executed.

[0144] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0145] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0146] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0147] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0148] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0149] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0150] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0151] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity 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, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0152] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0153] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for assessing exposure of hazardous waste resource products, characterized in that: The following steps are involved: Collect production process, application scenarios and physical and chemical testing data of hazardous waste resource products; Inputting the physical and chemical test data of the whole process into a toxicity prediction model to obtain toxicity data, and classifying the production process and the application scenario to obtain impact data; Obtaining an impact factor according to the correlation between the impact data and the exposure, and establishing a risk index according to the toxicity data and the exposure; constructing an exposure assessment model based on the toxicity data, the impact factor and the risk index, optimizing the parameters of the exposure assessment model, inputting the data to be assessed into the exposure assessment model, and outputting an assessment result; The method for constructing an exposure assessment model based on the toxicity data, the impact factor and the risk index comprises: The toxicity data, influencing factors and risk index are divided into training set, test set and validation set. The training set is used to learn the prediction pattern of toxicity data and influencing factors for risk index based on the Xgboost algorithm, and an exposure assessment model is established. The test set is used to optimize the hyperparameters of the exposure assessment model, and the accuracy of the exposure assessment model is verified by the validation set.

2. The method for assessing exposure of hazardous waste resource products according to claim 1, characterized in that: Methods for collecting production processes, application scenarios and physical and chemical testing data of hazardous waste resource products, including: The physical and chemical testing data of the whole process include the physical and chemical testing data of hazardous wastes, hazardous waste resource products, and by-products in the application scenarios of hazardous waste resource products. The physical and chemical testing data include the content of heavy metals and organic pollutants.

3. The method for assessing exposure of hazardous waste resource products according to claim 1, characterized in that: The method of inputting the whole process physical and chemical test data into a toxicity prediction model to obtain toxicity data comprises: The pre-trained toxicity prediction model obtains the median lethal dose, carcinogenic risk, and teratogenicity of pollutants through the pollutant dictionary and calculates the toxicity coefficient: Where T is the toxicity coefficient, LD50 is the median lethal dose, CP is the carcinogenic risk, J is the teratogenicity, w1, w2 and w3 are the weights of LD50, CP and J respectively, and the toxicity data include the toxicity coefficient and concentration of the pollutant.

4. The method for assessing exposure of hazardous waste resource products according to claim 1, characterized in that: The method for classifying the production process and the application scenario to obtain impact data includes: One-hot encoding is used to convert production processes and application scenarios into numerical vectors, and cluster centers are randomly selected to cluster production processes and application scenarios. The initial update formula of cluster centers is: where v i is the cluster center of cluster i, ω l and ω b are the weights of the lower approximation set and the boundary region, ω ll and ω lb is the weight of the lower approximation set and the boundary area after adding the imbalance adjustment parameter, x j is the jth data point, h ij is the membership of data point j to cluster i, m is the fuzzy coefficient, C i is cluster i, C i is the lower approximation set of cluster i, is the boundary area, is the upper approximation set of cluster i, is an empty set, The membership formula of the newly added data point to cluster i is: in, To add new data points The number of data points belonging to cluster i in the neighborhood data of yes The number of neighborhood data points within the neighborhood radius ζ, for and v i The spacing, Represents data points The local density of k is the number of clusters, The update formula of cluster center after adding new data points: in, for The number of data points in is the number of new data points.

5. The method for assessing exposure of hazardous waste resource products according to claim 1, characterized in that: The method for obtaining the impact factor according to the correlation between the impact data and the exposure degree includes: The exposure pathway is obtained based on the impact data, and the formula for calculating the exposure degree is: Where κ is the exposure, is the number of exposure pathways, is the amount of pollutant, For the Pollutants in the exposure pathway The concentration of For the The frequency of exposure through the exposure route, For the Duration of exposure for each exposure route, For the The exposure efficiency of each exposure route, For the Bioavailability by exposure route, For the The average exposure time for each exposure route, For the Pollutants by exposure pathway Environmental concentration, Using exposure as the target value, toxicity data and impact data as input vectors, the SVR model is trained. The Bayesian optimization algorithm is used to optimize the hyperparameters of the SVR model. The objective function of the SVR model is: Where f is the objective function, v is the weight vector of the SVR model, b is the bias term, M is the number of samples, and a i is the i-th input vector, e i is the i-th target value, ∈ is the threshold of insensitive loss, ∈∈(0,0.01), λ is the regularization coefficient, ρ is the weight of the regularization term, ρ∈[0,1], ||·|| represents the Euclidean distance, The Gaussian process used in Bayesian optimization is: in is the objective function at the hyperparameter point The value at , gp is the probability distribution of the Gaussian process, is the kernel function, and is a hyperparameter point in the hyperparameter search space, the Gaussian process mean is 0, and the kernel function is a Gaussian kernel function. Collection function: in is the predicted mean, is the prediction standard deviation, θ is the hyperparameter search space of the SVR model, β is the control coefficient of the control acquisition strategy, and the update rule of the prediction distribution is: in According to D 1:t and The objective function is obtained in The predicted distribution at The mean and variance are and Gaussian distribution, is the objective function in the tth iteration The prediction variance at 1:t is the set of hyperparameter points for the 1st to tth iterations, t is the current number of iterations, and f 1:t D 1:t The corresponding objective function Given a new hyperparameter point at the tth iteration The predicted mean at , is the expected value predicted using the Gaussian process, and K is the t hyperparameter points from the 1st to the tth iteration. The kernel matrix, K -1 is the inverse matrix of K, k is the kernel vector of the hyperparameter point at the t+1th iteration; Iterate the optimization until the objective function value is less than 0.05, calculate the SHAP value of the input vector for the target value as the correlation, and take the correlation of the influencing data as the influencing factor of the corresponding hazardous waste resource products.

6. The method for assessing exposure of hazardous waste resource products according to claim 1, characterized in that: A method for establishing a risk index based on the toxicity data and the exposure, comprising: The formula for calculating the risk index is: Among them, C 风险 is the risk index, is the amount of pollutant, For the The toxicity coefficient of the pollutant, κ is the exposure, For the The exposure concentration of the pollutant is the total hazard quotient, The non-carcinogenic hazards of these pollutants in all exposure pathways, including corrosiveness, neurotoxicity and ecotoxicity, For the The carcinogenic risk of a given exposure route is given by e, where e is the base of the natural logarithm.

7. The method for assessing exposure of hazardous waste resource products according to claim 1, characterized in that: The hyperparameters of the exposure assessment model are used as particles, and the mean square error of the exposure assessment model is used as the fitness value of the particles. The particles are ranked from small to large according to the fitness value, and the particle mass and inertial mass are calculated according to the ranking: Where m i (t) and m j (t) are the masses of the i-th particle and the j-th particle in the t-th iteration, respectively. i (t) is the rank of particle i at iteration t, N is the total number of particles, The formula for calculating the gravitational force between particles is: in is the gravitational force on the ith particle in the tth iteration, ω j is the weight of the jth particle, ranging from [0,1], is the gravitational force exerted on the i-th particle in the d-th dimension by the j-th particle in the t-th iteration, G(t) is the gravitational constant of the t-th iteration, G(t) = G0(t)·e -α ( t ) ·t / T , e is the base of the natural logarithm, α(t) is the attenuation coefficient of the tth iteration, T is the maximum number of iterations, M i (t) and M j (t) are the inertial masses of the ith particle and the jth particle in the tth iteration, R ij (t) is the Euclidean distance between the ith particle and the jth particle in the tth iteration, ε is the allowable error, and are the positions of the ith particle and the jth particle in the dth dimension in the tth iteration, Dynamically adjust the attenuation coefficient of the gravitational constant and the initial gravitational constant according to the number of iterations: Among them, α0 is the initial attenuation coefficient, β is the adjustment parameter, G0(t) is the initial gravitational constant of the tth iteration, G max and G min are the maximum and minimum values ​​of the initial gravitational constant, Particle velocity update formula and position update formula: in and are the velocities of the i-th particle in d dimension at the t-th iteration and the t+1-th iteration, respectively. is a random number between [0,1], is the acceleration of the ith particle in the d dimension at the tth iteration, f i (t) is the fitness value of particle i in the tth iteration, k v is the speed adjustment factor, f best (t) and f worst (t) are the minimum and maximum fitness values ​​in the tth iteration, is the position of the optimal solution in the dth dimension in the tth iteration. The optimal solution is the particle with the smallest fitness value. is the position of the i-th particle in the d-dimensional space at the t+1-th iteration. n(t) is a random noise term, which is a Gaussian noise with zero mean and time-varying variance. The optimal solution is output after the preset maximum number of iterations is reached.

Images