Data processing method and system for tailings plant remediation

Abstract

The present application provides a data processing method and system for tailing pond plant remediation, relating to the field of data processing, which comprises: obtaining a plurality of historical tailing pond plant remediation cases; determining a multi-dimensional remediation factor set based on the plurality of historical tailing pond plant remediation cases, wherein the multi-dimensional remediation factor set comprises a target tailing composition factor, a target soil state factor and a target meteorological factor; collecting remediation auxiliary features of a tailing pond to be remediated based on the multi-dimensional remediation factor set; searching for a reference historical tailing pond plant remediation case from the plurality of historical tailing pond plant remediation cases based on the remediation auxiliary features of the tailing pond to be remediated; determining a plant remediation strategy for the tailing pond to be remediated based on the reference historical tailing pond plant remediation case; collecting real-time remediation data of the tailing pond to be remediated during the execution of the plant remediation strategy of the tailing pond to be remediated, and performing remediation effect evaluation, which has the advantages of improving the efficiency and reliability of tailing pond plant remediation.

Description

Technical Field

[0001] This invention relates to the field of data processing, and in particular to a data processing method and system for phytoremediation of tailings ponds. Background Technology

[0002] Tailings ponds are artificial structures or natural reservoirs used to store tailings (solid waste remaining after ore beneficiation) during mining operations. They typically consist of tailings dams, drainage systems, and the pond area itself. As a concentrated source of environmental problems stemming from mining, tailings ponds commonly suffer from heavy metal pollution (such as lead, cadmium, and arsenic exceeding standards by several to dozens of times), along with deteriorating physicochemical properties (such as strong acidity / alkalinity, lack of organic matter, and structural fragmentation) and ecological degradation (such as loss of biodiversity and severe soil erosion). Traditional remediation technologies (such as topsoil covering and chemical stabilization) can reduce pollution risks in the short term, but they have limitations such as high costs, the risk of secondary pollution, and insufficient restoration of ecological functions. Against this backdrop, phytoremediation technology, with its advantages of low cost, environmental friendliness, and sustainability, has gradually become the mainstream direction for tailings pond ecological restoration. This technology achieves the dual goals of pollution control and ecological restoration through the absorption, fixation, or degradation of pollutants by hyperaccumulating plants, combined with soil amendment and the synergistic effect of microorganisms.

[0003] However, the effectiveness of phytoremediation is affected by multiple variables such as soil conditions, plant species, and climate factors. Traditional experience-driven remediation models are difficult to accurately match the needs of complex environments, and data-driven intelligent methods are urgently needed to improve remediation efficiency and reliability.

[0004] Therefore, there is a need to provide data processing methods and systems for phytoremediation of tailings ponds to improve the efficiency and reliability of phytoremediation. Summary of the Invention

[0005] This invention provides a data processing method for phytoremediation of tailings ponds, comprising: acquiring multiple historical tailings pond phytoremediation cases; determining a multidimensional remediation factor set based on the multiple historical tailings pond phytoremediation cases, wherein the multidimensional remediation factor set includes target tailings component factors, target soil state factors, and target meteorological factors; collecting remediation auxiliary features of the tailings pond to be remediated based on the multidimensional remediation factor set; searching for reference historical tailings pond phytoremediation cases from the multiple historical tailings pond phytoremediation cases based on the remediation auxiliary features of the tailings pond to be remediated; determining a phytoremediation strategy for the tailings pond to be remediated based on the reference historical tailings pond phytoremediation cases; and collecting real-time remediation data of the tailings pond to be remediated during the execution of the phytoremediation strategy for the tailings pond to be remediated, and evaluating the remediation effect.

[0006] Furthermore, the historical tailings dam phytoremediation cases include tailings composition characteristics, soil condition characteristics, meteorological characteristics, and phytoremediation strategies. Based on multiple historical tailings dam phytoremediation cases, a multidimensional remediation factor set is determined, including: determining multiple tailings composition factors, soil condition factors, and meteorological factors; screening multiple tailings composition factors, soil condition factors, and meteorological factors based on multiple historical tailings dam phytoremediation cases; and using an improved genetic algorithm, selecting target tailings composition factors, target soil condition factors, and target meteorological factors from the retained tailings composition factors, soil condition factors, and meteorological factors after screening, based on multiple historical tailings dam phytoremediation cases.

[0007] Furthermore, based on multiple historical tailings dam phytoremediation cases, several tailings component factors, soil condition factors, and meteorological factors were screened, including: for each tailings component factor, determining the difference between any two historical tailings dam phytoremediation cases and the text similarity of the phytoremediation strategies in any two historical tailings dam phytoremediation cases, and substituting these differences and text similarity as two variables into the correlation coefficient calculation formula to obtain the correlation coefficient between the tailings component factor and the phytoremediation strategy; for each soil condition factor, determining the difference between any two historical tailings dam phytoremediation cases and the text similarity of the phytoremediation strategies in any two historical tailings dam phytoremediation cases, and substituting these differences and text similarity as two variables into the correlation coefficient calculation formula to obtain the correlation coefficient between the tailings component factor and the phytoremediation strategy; and for each soil condition factor, determining the difference between any two historical tailings dam phytoremediation cases and the text similarity of the phytoremediation strategies in any two historical tailings dam phytoremediation cases, and substituting these differences and text similarity as two variables into the correlation coefficient calculation formula to obtain the correlation coefficient between the tailings component factor and the phytoremediation strategy. The soil condition factor is calculated by substituting the difference between any two historical tailings dam phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings dam phytoremediation cases into the correlation coefficient calculation formula. For each meteorological factor, the difference between any two historical tailings dam phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings dam phytoremediation cases are determined. These two factors are then used as variables in the correlation coefficient calculation formula to obtain the correlation coefficient between the meteorological factor and the phytoremediation strategy. Tailings component factors, soil condition factors, and meteorological factors with absolute correlation coefficients greater than the absolute value threshold are retained.

[0008] Furthermore, using an improved genetic algorithm, based on multiple historical tailings dam phytoremediation cases, target tailings component factors, target soil state factors, and target meteorological factors are screened from the retained tailings component factors, soil state factors, and meteorological factors. This includes: generating multiple individuals based on the correlation coefficients between tailings component factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies. Specifically, when generating each individual, the absolute values ​​of the correlation coefficients between tailings component factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies are used to sample the tailings component factors, soil state factors and meteorological factors. The individual is a combination of the sampled tailings component factors, soil state factors and meteorological factors. A fitness function is constructed. Based on the fitness function and the correlation coefficients between tailings component factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, multiple individuals are optimized to determine the target tailings component factors, target soil state factors, and target meteorological factors.

[0009] Furthermore, the dependent variable of the fitness function includes the correlation coefficient between the individual and the phytoremediation strategy. Specifically, for each individual, a feature vector corresponding to the historical tailings pond phytoremediation case is constructed based on the tailings component factor, soil state factor, and meteorological factor included in the individual. The feature vector consists of the specific values ​​of the tailings component factor, soil state factor, and meteorological factor included in the individual corresponding to the historical tailings pond phytoremediation case. For any two historical tailings pond phytoremediation cases, the similarity of the feature vectors of the individuals corresponding to the two historical tailings pond phytoremediation cases is calculated. The similarity of the feature vectors of the individuals corresponding to the two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategy are used as two variables and substituted into the formula for calculating the correlation coefficient to obtain the correlation coefficient between the individual and the phytoremediation strategy. The larger the absolute value of the correlation coefficient between the individual and the phytoremediation strategy, the larger the value of the fitness function.

[0010] Furthermore, based on the fitness function and the correlation coefficients between tailings component factors and phytoremediation strategies, soil condition factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, multiple individuals are optimized. This includes: using a roulette wheel selection algorithm to select retained individuals based on their fitness; and guiding individuals to undergo crossover and mutation operations based on the correlation coefficients between tailings component factors and phytoremediation strategies, soil condition factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies to optimize multiple individuals. Specifically, in the crossover operation, for any two selected retained individuals, the tailings component factor with the largest difference in correlation coefficient with the phytoremediation strategy, the soil condition factor with the largest difference in correlation coefficient with the phytoremediation strategy, and the meteorological factor with the largest difference in correlation coefficient with the phytoremediation strategy are swapped. In the mutation operation, for each selected retained individual after the swap operation, the tailings component factor, soil condition factor, and meteorological factor with the smallest absolute value of correlation coefficient are replaced with tailings component factors, soil condition factors, and meteorological factors with larger absolute values ​​of correlation coefficients that were not selected by the individual.

[0011] Furthermore, based on the auxiliary features of the tailings dam to be restored, reference historical tailings vegetative restoration cases are sought from multiple historical tailings dam vegetative restoration cases. This includes: for each historical tailings dam vegetative restoration case, determining the auxiliary features of the historical tailings dam vegetative restoration case based on a multidimensional set of restoration factors; calculating the similarity between the auxiliary features of the tailings dam to be restored and the auxiliary features of the historical tailings dam vegetative restoration cases; and based on the similarity between the auxiliary features of the tailings dam to be restored and the auxiliary features of the historical tailings dam vegetative restoration cases, finding reference historical tailings dam vegetative restoration cases from multiple historical tailings dam vegetative restoration cases.

[0012] Furthermore, real-time remediation data of the tailings dam to be remediated is collected, including: identifying multiple monitoring points of the tailings dam to be remediated; collecting soil condition data of multiple monitoring points of the tailings dam to be remediated at multiple remediation time points; identifying multiple target monitoring points based on the soil condition data of multiple monitoring points of the tailings dam to be remediated collected at multiple remediation time points; and collecting real-time remediation data of the tailings dam to be remediated based on the multiple target monitoring points.

[0013] Furthermore, based on soil state data from multiple monitoring points of the tailings dam to be repaired collected at multiple remediation time points, multiple target monitoring points are determined, including: for any two monitoring points, calculating the cosine similarity and Pearson correlation coefficient of the soil state data of the tailings dam to be repaired at multiple remediation time points for the two monitoring points; based on the cosine similarity and Pearson correlation coefficient of the soil state data of the tailings dam to be repaired at multiple remediation time points for the two monitoring points, multiple target monitoring points are determined. Specifically, if the cosine similarity of the soil state data of the tailings dam to be repaired at multiple remediation time points for the two monitoring points is greater than the cosine similarity threshold or the Pearson correlation coefficient is greater than the Pearson correlation coefficient threshold, then one of the monitoring points is retained as the target monitoring point.

[0014] This invention provides a data processing system for phytoremediation of tailings ponds, applying the aforementioned data processing method for phytoremediation of tailings ponds, comprising: a case acquisition module for acquiring multiple historical tailings pond phytoremediation cases; a factor determination module for determining a multidimensional remediation factor set based on the multiple historical tailings pond phytoremediation cases, wherein the multidimensional remediation factor set includes target tailings component factors, target soil state factors, and target meteorological factors; a data acquisition module for acquiring remediation auxiliary data of the tailings pond to be remediated based on the multidimensional remediation factor set; a case search module for searching for reference historical tailings pond phytoremediation cases from the multiple historical tailings pond phytoremediation cases based on the remediation auxiliary data of the tailings pond to be remediated; a strategy determination module for determining a phytoremediation strategy for the tailings pond to be remediated based on the reference historical tailings pond phytoremediation cases; and a remediation evaluation module for acquiring real-time remediation data of the tailings pond to be remediated during the execution of the phytoremediation strategy and evaluating the remediation effect.

[0015] Compared with existing technologies, the data processing method and system for phytoremediation of tailings ponds provided by this invention have at least the following beneficial effects:

[0016] 1. A multi-dimensional set of remediation factors (covering tailings composition, soil condition, and meteorological conditions) extracted from historical cases is used to accurately screen key factors by quantifying the correlation between factors and remediation strategies, avoiding interference from irrelevant information, and matching the most suitable reference cases for tailings dams to be remediated, thereby improving the scientific nature and pertinence of remediation strategies. 2. Real-time remediation data is dynamically collected and the layout of monitoring points is optimized. By calculating the state similarity and correlation coefficient between monitoring points, representative target monitoring points are selected, which not only comprehensively captures the spatial heterogeneity of the remediation process but also reduces data collection costs, enabling dynamic evaluation of remediation effects and timely adjustment of strategies. 3. The entire process replaces experience-based judgment with objective data, forming a closed-loop management system from strategy formulation to effect feedback. For example, remediation bottlenecks (such as substandard soil pH) can be traced based on real-time data and precise regulation can be triggered, significantly improving remediation efficiency, success rate, and economy, providing a standardized and replicable intelligent solution for the ecological remediation of tailings dams in complex environments.

[0017] 2. Target remediation factors are screened from a massive pool of factors using an improved genetic algorithm, constructing a multidimensional remediation factor set covering tailings composition, soil conditions, and meteorological conditions. Traditional methods often rely on single factors or empirical selection, easily overlooking the interactions between factors (e.g., high pH values ​​reduce heavy metal activity and affect plant selection). However, by calculating the correlation coefficients between factors and remediation strategies, combined with the global search capability of the genetic algorithm, key factors that significantly affect remediation effectiveness are accurately identified.

[0018] 3. Traditional assessment methods typically use fixed monitoring points, which can lead to high data collection costs and insufficient data representativeness in some areas (such as large dynamic differences in remediation between highly polluted areas and background areas). This invention introduces a dynamic monitoring point screening mechanism in the remediation effect assessment stage. By calculating the state similarity and correlation coefficient between monitoring points, target monitoring points with strong representativeness and low information redundancy are selected from the initially deployed monitoring points. Attached Figure Description

[0019] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0020] Figure 1 This is a flowchart illustrating a data processing method for phytoremediation of tailings ponds, according to some embodiments of this specification.

[0021] Figure 2 This is a schematic diagram of a data processing system for phytoremediation of tailings ponds, as shown in some embodiments of this specification. Detailed Implementation

[0022] To more clearly illustrate the technical solutions of the embodiments in this specification, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some examples or embodiments of this specification. For those skilled in the art, these drawings can be applied to other similar scenarios without creative effort. Unless obvious from the context or otherwise specified, the same reference numerals in the drawings represent the same structures or operations.

[0023] Figure 1 This is a flowchart illustrating a data processing method for phytoremediation of tailings ponds, as shown in some embodiments of this specification. Figure 1 As shown, the data processing method for phytoremediation of tailings ponds may include the following steps.

[0024] Step 110: Obtain multiple historical cases of phytoremediation of tailings ponds.

[0025] Among them, historical tailings dam phytoremediation cases include tailings composition characteristics, soil condition characteristics, meteorological characteristics, and phytoremediation strategies.

[0026] Specifically, the characteristics of tailings composition can include the types of heavy metals (such as lead, cadmium, zinc, copper, etc.) and their content range, as well as the physicochemical properties of tailings such as mineral composition (such as sulfides, oxides, silicates, etc.), pH value, and organic matter content.

[0027] Soil condition characteristics can include physical properties such as soil texture (e.g., sandy soil, clay soil, loam), structure (e.g., granular structure, blocky structure), porosity, and water content, as well as biochemical properties such as soil nutrient content (e.g., nitrogen, phosphorus, potassium), enzyme activity, and microbial community structure.

[0028] Meteorological characteristics may include average annual precipitation, average temperature, average sunshine duration, and average wind speed.

[0029] Phytoremediation strategies may include the plant species used (such as hyperaccumulators and heavy metal tolerant plants), planting patterns (monoculture, mixed planting, intercropping), planting density, fertilization programs (fertilizer type, dosage, and application frequency), irrigation management (irrigation volume, frequency, and water source), and auxiliary measures (such as EDTA chelating agent addition and microbial inoculation).

[0030] In historical cases of phytoremediation of tailings ponds, the phytoremediation strategy can be determined based on expert experience or experimental data.

[0031] Step 120: Based on multiple historical cases of phytoremediation of tailings ponds, determine a multidimensional set of remediation factors.

[0032] The multidimensional remediation factor set includes target tailings component factors, target soil state factors, and target meteorological factors.

[0033] In some embodiments, step 120 specifically includes:

[0034] Multiple tailings composition factors, soil condition factors, and meteorological factors were determined.

[0035] Based on multiple historical cases of phytoremediation of tailings ponds, multiple tailings component factors, soil state factors, and meteorological factors were screened.

[0036] Using an improved genetic algorithm, and based on multiple historical tailings dam phytoremediation cases, target tailings components, target soil conditions, and target meteorological factors were selected from the retained tailings components, soil conditions, and meteorological factors.

[0037] Specifically, tailings composition factors refer to quantitative indicators that reflect the characteristics of heavy metal pollution and mineral composition of tailings, such as heavy metal content (e.g., lead and cadmium concentrations), heavy metal forms (percentage of exchangeable forms), mineral types (sulfide / oxide ratio), and pH value.

[0038] Soil state factors are used to describe the physical, chemical, and biological properties of soil, such as soil texture (sand / clay ratio), structure (integrity of aggregate structure), porosity, and nutrient content (available nitrogen / phosphorus / potassium).

[0039] Meteorological factors characterize the climatic conditions of tailings ponds, such as average annual precipitation, average temperature, average sunshine duration, and average wind speed.

[0040] In some embodiments, based on multiple historical tailings dam phytoremediation cases, multiple tailings component factors, soil state factors, and meteorological factors are screened, including:

[0041] For each tailings component factor, the difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are determined. The difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are taken as two variables and substituted into the correlation coefficient calculation formula to obtain the correlation coefficient between the tailings component factor and the phytoremediation strategy.

[0042] For each soil state factor, the difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are determined. The difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are taken as two variables and substituted into the correlation coefficient calculation formula to obtain the correlation coefficient between soil state factors and phytoremediation strategies.

[0043] For each meteorological factor, the difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are determined. The difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are taken as two variables and substituted into the correlation coefficient calculation formula to obtain the correlation coefficient between the meteorological factor and the phytoremediation strategy.

[0044] Tailings composition factors, soil state factors, and meteorological factors with correlation coefficients greater than an absolute value threshold (e.g., 0.7) are retained.

[0045] Specifically, by quantifying the correlation between various factors and phytoremediation strategies, factors that significantly influence remediation outcomes are identified. For each tailings component factor, the absolute value of the difference between any two historical cases is first calculated. This difference reflects the variation of that tailings component factor among different historical tailings pond phytoremediation cases.

[0046] Meanwhile, by comparing the similarity (e.g., text similarity based on strategy descriptions or expert scores) of the phytoremediation strategies (such as plant species combinations, fertilization methods, auxiliary measures, etc.) adopted in these two historical tailings pond phytoremediation cases, the strategy differences are quantified. Specifically, a method based on term frequency-inverse document frequency (TF-IDF) and cosine similarity can be used to calculate the text similarity of the phytoremediation strategies adopted in two historical tailings pond phytoremediation cases. First, preprocess the phytoremediation strategies adopted in the two historical tailings pond phytoremediation cases, including removing stop words (such as meaningless words like "of", "and", "is", etc.), punctuation marks, and performing word segmentation to convert the text into a set of words. Then, use the TF-IDF algorithm to assign weights to the words in each phytoremediation strategy text. TF (term frequency) reflects the frequency of a word appearing in the phytoremediation strategy text, and IDF (inverse document frequency) measures the general importance of a word. The combination of the two can highlight the weights of words that are important in a specific text and rare in other texts. Next, represent the two phytoremediation strategy texts as vectors, where each dimension of the vector corresponds to a word, and the value of the dimension is the TF-IDF weight of the word in the text. After that, use the cosine similarity formula to calculate the similarity of these two vectors. Cosine similarity measures the similarity between two vectors by calculating the cosine value of the angle between them. The value range is between [-1, 1]. The closer the value is to 1, the more similar the semantics of the two texts; the closer it is to -1, the greater the difference; and close to 0 indicates that the two are basically semantically unrelated. Through this method, the similarity of the phytoremediation strategy texts adopted in two historical tailings pond phytoremediation cases can be quantitatively evaluated.

[0047] Take the difference in tailings component factors corresponding to any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases as two variables, and substitute them into the calculation formula of the correlation coefficient (e.g., Pearson correlation coefficient, distance correlation coefficient, etc.) to obtain the correlation coefficient between the tailings component factors and the phytoremediation strategies. Tailings component factors with an absolute value of the correlation coefficient greater than the absolute value threshold (e.g., 0.5) can be retained.

[0048] As an example, according to the calculation formula of the Pearson correlation coefficient, the correlation coefficient between the tailings component factors and the phytoremediation strategies can be obtained:

[0049]

[0050] where is the difference in tailings component factors corresponding to the i-th case pair, and each case pair includes two historical tailings pond phytoremediation cases, is the text similarity of the phytoremediation strategies of the i-th case pair, and n is the total number of case pairs.

[0051] There are several ways to determine the absolute value threshold. First, based on rules of thumb, referencing thresholds used in previous similar studies in related fields. Many predecessors have accumulated experience in determining thresholds for correlation coefficient screening, which can be used as a reference. Second, statistical methods can be employed, calculating a relatively reasonable absolute value threshold based on the data distribution characteristics, such as standard deviation and variance. For example, a method based on ranking proportions can be used to determine the absolute value threshold. First, calculate the absolute values ​​of the correlation coefficients between all tailings component factors and related indicators such as phytoremediation strategies, and arrange them in descending order. This step clearly shows the distribution of the correlation degree of each factor. Next, determine the retention proportion based on research needs and sample size. For example, if you want to focus on factors with strong correlations, you can set the retention target to the top 30% of tailings component factors. Multiply the sample size by this proportion to determine the number of factors to retain. Then, find the absolute value of the correlation coefficient at the corresponding position after ranking; this value is the determined absolute value threshold. For example, if there are 100 tailings component factors, and the top 30% (30 factors) are retained, then the absolute value of the correlation coefficient of the 30th factor is the threshold. The advantage of this method is that it incorporates the actual distribution of the data, avoiding the arbitrariness of subjectively setting the threshold. Furthermore, the retention ratio can be flexibly adjusted according to the research focus. To more comprehensively explore potential correlation factors, the retention ratio can be appropriately increased; if the focus is on key factors, the retention ratio can be decreased, thus scientifically and rationally determining the absolute value threshold.

[0052] The methods for calculating the correlation coefficients between soil state factors and phytoremediation strategies, the correlation coefficients between meteorological factors and phytoremediation strategies, and the methods for screening soil state factors and meteorological factors are similar to the methods for calculating the correlation coefficients between tailings component factors and phytoremediation strategies and the methods for screening tailings component factors, and will not be elaborated here.

[0053] By calculating the correlation coefficient between the factor difference between any two cases and the text similarity of the phytoremediation strategy, complex relationships can be captured (e.g., when the cadmium content varies greatly in a tailings pond, the strategy always uses hyperaccumulating plants, indicating that cadmium content is a key driving factor in strategy formulation), avoiding the omission of important variables. Secondly, based on the difference-similarity analysis of multiple historical tailings pond phytoremediation cases, redundant factors (such as factors that are weakly correlated with the strategy or only indirectly affected by other factors) can be systematically eliminated, reducing the computational complexity of subsequent optimization algorithms (e.g., reducing the number of candidate factors from 100 to 20, reducing the computational load by 80%).

[0054] In some embodiments, using an improved genetic algorithm, based on multiple historical tailings dam phytoremediation cases, target tailings component factors, target soil state factors, and target meteorological factors are screened from the pre-screened tailings component factors, soil state factors, and meteorological factors, including:

[0055] Based on the correlation coefficients between screened tailings component factors and phytoremediation strategies, soil condition factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, multiple individuals are generated. Specifically, when generating each individual, the absolute values ​​of the correlation coefficients between the screened tailings component factors, soil condition factors and phytoremediation strategies, and meteorological factors are used to sample the tailings component factors, soil condition factors and meteorological factors. Each individual is a combination of the sampled tailings component factors, soil condition factors and meteorological factors. A sampling method based on the absolute value of the correlation coefficient is used when generating each individual. For tailings component factors, each tailings component factor is assigned a corresponding sampling weight according to the magnitude of its absolute correlation coefficient with the phytoremediation strategy; the larger the absolute value, the higher the probability that the factor will be selected. For example, if a tailings component factor has the highest absolute value of its correlation coefficient with phytoremediation strategies among all tailings component factors, then this factor is significantly more likely to be selected during sampling than other factors with smaller absolute values ​​of correlation coefficients. Similarly, for soil condition factors and meteorological factors, the sampling weight is determined according to the absolute value of their respective correlation coefficients with phytoremediation strategies. During the sampling process, sampling operations are performed independently from the three categories of factors: tailings component factors, soil condition factors, and meteorological factors. Ultimately, each individual is composed of a combination of factors sampled from these three categories. In other words, an individual contains several tailings component factors, several soil condition factors, and several meteorological factors. This method of generating an individual set highlights factors closely related to phytoremediation strategies, allowing subsequent analyses and studies based on these individuals to focus more on factors that significantly influence phytoremediation strategies, thus improving efficiency. As an example, the corresponding sampling weight of tailings component factors can be calculated using the following formula:

[0056]

[0057] in, Let be the sampling weight corresponding to the i-th tailings component factor. Let be the correlation coefficient between the i-th tailings component factor and the phytoremediation strategy. Let be the correlation coefficient between the i-th tailings component factor and the phytoremediation strategy. This represents the total number of remaining tailings component factors after screening.

[0058] Construct a fitness function, where the dependent variable of the fitness function includes the correlation coefficient between the individual and the phytoremediation strategy;

[0059] Based on the fitness function and the correlation coefficients between tailings composition factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, multiple individuals are optimized to determine the target tailings composition factors, target soil state factors, and target meteorological factors.

[0060] In some embodiments, optimization is performed on multiple individuals based on the correlation coefficients between fitness functions and tailings component factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, including:

[0061] The roulette wheel selection algorithm is used to select and retain individuals based on their fitness.

[0062] Based on the correlation coefficients between tailings composition factors and phytoremediation strategies, soil condition factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, individuals are guided to perform cross-operations and variation operations to optimize multiple individuals.

[0063] Specifically, each individual may include at least one tailings component factor, soil condition factor, and meteorological factor.

[0064] In the initial population generation phase, the algorithm does not randomly select factor combinations. Instead, it performs weighted sampling based on the correlation coefficients between tailings component factors and phytoremediation strategies, soil condition factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies. For example, if the correlation coefficient of a certain tailings component factor (such as the exchangeable content of cadmium) is 0.8, while the correlation coefficient of another tailings component factor (such as total zinc) is 0.3, the former has a significantly higher probability of being selected as an individual. This design ensures that the initial population contains more factors strongly correlated with the strategy, accelerating convergence to the optimal solution.

[0065] For each individual, a feature vector is constructed based on the tailings composition factors, soil state factors, and meteorological factors included in the individual. The feature vector consists of the specific values ​​of the tailings composition factors, soil state factors, and meteorological factors corresponding to the historical tailings dam phytoremediation cases. For any two historical tailings dam phytoremediation cases, the similarity of the feature vectors of the individuals corresponding to the two historical tailings dam phytoremediation cases is calculated. The similarity of the feature vectors of the individuals corresponding to the two historical tailings dam phytoremediation cases and the textual similarity of the phytoremediation strategies are used as two variables and substituted into the formula for calculating correlation coefficients (e.g., Pearson correlation coefficient, distance correlation coefficient, etc.) to obtain the correlation coefficient between the individual and the phytoremediation strategy.

[0066] As an example, the correlation coefficient between an individual and a phytoremediation strategy can be obtained using the formula for calculating the Pearson correlation coefficient:

[0067]

[0068] in, Let be the similarity of the feature vectors of the corresponding individuals in the i-th case pair. Each case pair includes two historical tailings pond phytoremediation cases. Let be the text similarity of the phytoremediation strategy for the i-th case pair, and n be the total number of case pairs.

[0069] The dependent variable of the fitness function includes the correlation coefficient between the individual and the phytoremediation strategy. The larger the absolute value of the correlation coefficient between the individual and the phytoremediation strategy, the larger the value of the fitness function.

[0070] As an example only, the fitness function can be:

[0071]

[0072] in, For the fitness function, This represents the absolute value of the correlation coefficient between the individual and the phytoremediation strategy. It is an optional positive constant used to adjust the range of fitness values, for example, .

[0073] Traditional genetic algorithms randomly select segments of parent individuals for crossover. However, improved genetic algorithms guide crossover point selection based on the correlation coefficients between tailings components, soil conditions, and meteorological factors with the phytoremediation strategy. For example, in the crossover operation, for any two retained individuals, the tailings components with the largest difference in correlation coefficient with the phytoremediation strategy, the soil conditions with the largest difference in correlation coefficient with the phytoremediation strategy, and the meteorological factors with the largest difference in correlation coefficient with the phytoremediation strategy are exchanged.

[0074] Traditional mutation randomly flips a gene locus, while improved genetic algorithms prioritize mutation of factors with lower correlation coefficients. For example, in the mutation operation, for each individual that remains after the exchange operation, the tailings component factor, soil state factor, and meteorological factor with the smallest absolute value of the correlation coefficient in that individual are replaced with tailings component factors, soil state factors, and meteorological factors with larger absolute values ​​of the correlation coefficient that were not selected by the individual.

[0075] Optimization termination conditions may include: reaching the maximum number of iterations, or the fitness function value not significantly improving for multiple consecutive generations.

[0076] In the initial population generation stage, correlation coefficient-weighted sampling is used to prioritize highly correlated factors (such as the content of cadmium exchangeable states), ensuring the quality of the initial population and accelerating convergence to the optimal solution, improving efficiency by approximately 40% compared to traditional random sampling. During individual optimization, a fitness function based on the correlation coefficient between factor combinations and strategies is constructed to directly quantify the contribution of the factor set to the remediation strategy, avoiding subjective weight allocation bias and making the selection results more scientific and interpretable. Crossover operations, guided by factor correlation coefficients, exchange low-correlation factors with high-correlation factors, effectively preventing local optima traps; mutation operations prioritize adjusting low-correlation factors, enhancing global search capabilities.

[0077] Step 130: Based on the multidimensional repair factor set, collect the repair auxiliary features of the tailings dam to be repaired.

[0078] Specifically, the auxiliary features for the remediation of tailings dams to be remediated may include the values ​​of target tailings component factors, target soil state factors, and target meteorological factors for each target tailings dam.

[0079] Step 140: Based on the auxiliary features of the tailings dam to be restored, referencing historical tailings vegetation restoration cases are found from multiple historical tailings dam vegetation restoration cases.

[0080] Specifically, it includes:

[0081] For each historical tailings dam vegetation remediation case, the remediation auxiliary characteristics of the historical tailings dam vegetation remediation case are determined based on a multidimensional remediation factor set;

[0082] Calculate the similarity between the auxiliary remediation features of the tailings dam to be remediated and the auxiliary remediation features of historical tailings dam vegetation remediation cases;

[0083] Based on the similarity between the auxiliary features of the tailings dam to be restored and the auxiliary features of historical tailings dam vegetation restoration cases, reference historical tailings dam vegetation restoration cases were found from multiple historical tailings dam vegetation restoration cases.

[0084] Specifically, similarity calculation algorithms (such as cosine similarity, Euclidean distance, or Mahalanobis distance) are used to quantify the similarity between the auxiliary features of the tailings dam to be restored and the auxiliary features of historical tailings dam vegetative restoration cases. Historical tailings dam vegetative restoration cases with a similarity greater than a similarity threshold (e.g., 60%) can be used as reference historical tailings dam vegetative restoration cases.

[0085] Step 150: Based on referenced historical cases of phytoremediation of tailings ponds, determine the phytoremediation strategy for the tailings pond to be remediated.

[0086] As an example only, historical tailings pond phytoremediation cases with the highest similarity can be referenced.

[0087] For another example, a strategy determination model can be used to determine the phytoremediation strategy for a tailings pond to be remediated based on reference historical tailings pond phytoremediation cases. The strategy determination model includes a feature encoding layer, a strategy inference layer, and an output layer: the feature encoding layer is used to extract feature vectors from historical tailings pond phytoremediation cases; the strategy inference layer adopts a Transformer architecture, whose self-attention mechanism can dynamically capture complex interactions between features (such as the synergistic effect of soil organic matter and microbial community on phytoremediation efficiency), and further mines higher-order nonlinear relationships through a multilayer perceptron (MLP); the output layer designs a dedicated decoder for different strategy types. For example, for classification tasks (such as plant species selection), a Softmax activation function is used to generate a probability distribution, and for regression tasks (such as fertilizer application prediction), a linear activation function is used to output continuous values.

[0088] Based on the concepts of transfer learning and few-shot learning, this paper addresses the scarcity of historical case data through a pre-training-fine-tuning paradigm: First, unsupervised pre-training is performed on large-scale environmental remediation datasets (such as publicly available databases of soil heavy metal pollution remediation) to learn general feature representation patterns (e.g., the correlation between heavy metal speciation and plant uptake rates). Then, supervised fine-tuning is performed on historical tailings pond phytoremediation cases to adapt the model to the feature distribution of specific scenarios (e.g., the remediation needs of acidic tailings in a certain region). During training, contrastive learning is employed to enhance feature discriminative power by maximizing the inner product of feature vectors from similar cases and minimizing the inner product of feature vectors from dissimilar cases, thereby improving the model's sensitivity to key remediation factors. Simultaneously, a multi-task learning framework is introduced to jointly optimize the predictive losses of multiple sub-strategies such as plant selection, soil improvement, and irrigation schemes, preventing the model from favoring a single strategy and neglecting the overall remediation effect.

[0089] In the data preprocessing stage, the features of historical cases are normalized and K-means clustering is used to generate case similarity maps to help the model learn the implicit relationships between cases. During model training, an adaptive learning rate optimizer (such as RAdam) is used, combined with an early stopping mechanism to prevent overfitting. Training is terminated when the validation set loss does not decrease for 10 consecutive rounds.

[0090] Step 160: During the implementation of the phytoremediation strategy for the tailings dam to be repaired, real-time remediation data of the tailings dam to be repaired is collected to evaluate the remediation effect.

[0091] Specifically, it includes:

[0092] Multiple monitoring points were identified for the tailings dam to be repaired;

[0093] Soil condition data were collected from multiple monitoring points at various remediation time points in the tailings dam to be remediated.

[0094] Based on soil condition data from multiple monitoring points of the tailings dam to be repaired collected at multiple remediation time points, multiple target monitoring points were identified.

[0095] Real-time repair data of the tailings dam to be repaired is collected based on multiple target monitoring points.

[0096] In some embodiments, multiple target monitoring points are determined based on soil state data from multiple monitoring points of the tailings dam to be remediated collected at multiple remediation time points, including:

[0097] For any two monitoring points, calculate the cosine similarity and Pearson correlation coefficient of the soil state data of the tailings dam to be repaired at multiple repair time points for the two monitoring points;

[0098] Based on the cosine similarity and Pearson correlation coefficient of the soil state data of the tailings pond to be repaired at multiple remediation time points of two monitoring points, multiple target monitoring points are determined. Specifically, if the cosine similarity of the soil state data of the tailings pond to be repaired at multiple remediation time points of two monitoring points is greater than the cosine similarity threshold (e.g., 70%) or the Pearson correlation coefficient is greater than the Pearson correlation coefficient threshold (e.g., 0.7), then one of the monitoring points is retained as the target monitoring point.

[0099] Specifically, the synchronicity of the remediation process is quantified by calculating the cosine similarity of soil state data of two monitoring points at multiple remediation time points. For example, if the heavy metal active content change curves of monitoring points A and B highly overlap at three consecutive remediation time points (cosine similarity > 90%), it indicates that their remediation responses are consistent. On the other hand, the correlation strength between the remediation effects of monitoring points is assessed by calculating state correlation coefficients (such as Pearson correlation coefficient or Spearman rank correlation coefficient). For example, the plant biomass of monitoring point C and the soil organic matter content of monitoring point D are significantly positively correlated (r > 0.8), indicating that there is an ecological coupling relationship between the two.

[0100] After the screening is completed, the remaining monitoring points will be used as target monitoring points.

[0101] Based on real-time remediation data collected from target monitoring points, including values ​​corresponding to target tailings component factors, target soil state factors, and target meteorological factors, a remediation effectiveness evaluation index system (including heavy metal removal rate, plant survival rate, and soil ecological function recovery degree) is constructed. This system, combined with machine learning models (such as random forests or LSTM neural networks), predicts remediation trends and dynamically adjusts remediation strategies (such as adding amendments or replanting hyperaccumulating plants in areas with slow heavy metal removal). For example, if target monitoring point data shows that the rate of decline in cadmium active content in a certain area is lower than expected, the model can trace this back to substandard soil pH or insufficient abundance of functional bacteria, thereby triggering precise control measures. This forms a closed-loop management system of "monitoring-evaluation-optimization," significantly improving the efficiency and reliability of phytoremediation of tailings ponds.

[0102] Figure 2 This is a schematic diagram of a data processing system for phytoremediation of tailings ponds, as shown in some embodiments of this specification. Figure 2 As shown, the data processing system for phytoremediation of tailings ponds may include a case acquisition module, a factor determination module, a data collection module, a case search module, a strategy determination module, and a remediation assessment module.

[0103] The case acquisition module is used to acquire multiple historical cases of phytoremediation of tailings ponds;

[0104] The factor determination module is used to determine a multidimensional set of remediation factors based on multiple historical tailings dam phytoremediation cases. The multidimensional set of remediation factors includes target tailings component factors, target soil state factors, and target meteorological factors.

[0105] The data acquisition module is used to collect auxiliary data for the repair of tailings dams to be repaired based on a multidimensional set of repair factors.

[0106] The case search module is used to find reference historical tailings vegetation remediation cases from multiple historical tailings dam vegetation remediation cases based on the remediation auxiliary data of the tailings dam to be remediated.

[0107] The strategy determination module is used to determine the phytoremediation strategy for the tailings dam to be remediated based on referenced historical tailings dam phytoremediation cases.

[0108] The remediation assessment module is used to collect real-time remediation data of the tailings dam to be remediated during the execution of the phytoremediation strategy and to evaluate the remediation effect.

[0109] The data processing system for phytoremediation of tailings ponds can be used to execute data processing methods for phytoremediation of tailings ponds, which will not be elaborated here.

[0110] Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be illustrative rather than limiting, and should be considered consistent with the teachings of this specification. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A data processing method for phytoremediation of tailings ponds, characterized in that, include: Acquire multiple historical cases of phytoremediation of tailings ponds; Based on multiple historical cases of phytoremediation of tailings ponds, a multidimensional set of remediation factors was determined, which includes target tailings component factors, target soil state factors, and target meteorological factors. Based on a multidimensional set of repair factors, auxiliary features for the repair of tailings dams to be repaired are collected. Based on the auxiliary characteristics of the tailings dam to be repaired, reference historical tailings vegetation restoration cases were found from multiple historical tailings dam vegetation restoration cases. Based on reference to historical cases of phytoremediation of tailings ponds, a phytoremediation strategy for the tailings pond to be remediated was determined. During the implementation of the phytoremediation strategy for the tailings dam to be repaired, real-time remediation data of the tailings dam to be repaired is collected to evaluate the remediation effect; Among them, the historical tailings pond phytoremediation cases include tailings composition characteristics, soil condition characteristics, meteorological characteristics, and phytoremediation strategies; Based on multiple historical cases of phytoremediation of tailings ponds, a multidimensional set of remediation factors was identified, including: Multiple tailings composition factors, soil condition factors, and meteorological factors were determined. Based on multiple historical cases of phytoremediation of tailings ponds, multiple tailings component factors, soil state factors, and meteorological factors were screened. Using an improved genetic algorithm, based on multiple historical tailings dam phytoremediation cases, target tailings component factors, target soil state factors, and target meteorological factors were screened from the retained tailings component factors, soil state factors, and meteorological factors. Based on multiple historical tailings dam phytoremediation cases, various tailings component factors, soil condition factors, and meteorological factors were screened, including: For each tailings component factor, the difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are determined. The difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are taken as two variables and substituted into the correlation coefficient calculation formula to obtain the correlation coefficient between the tailings component factor and the phytoremediation strategy. For each soil state factor, the difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are determined. The difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are taken as two variables and substituted into the correlation coefficient calculation formula to obtain the correlation coefficient between soil state factors and phytoremediation strategies. For each meteorological factor, the difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are determined. The difference between any two historical tailings pond phytoremediation cases and the text similarity of the phytoremediation strategies of any two historical tailings pond phytoremediation cases are taken as two variables and substituted into the correlation coefficient calculation formula to obtain the correlation coefficient between the meteorological factor and the phytoremediation strategy. Tailings component factors, soil state factors, and meteorological factors whose absolute values ​​of correlation coefficients are greater than the absolute value threshold are retained; Using an improved genetic algorithm, based on multiple historical tailings dam phytoremediation cases, target tailings component factors, target soil condition factors, and target meteorological factors were screened from the retained tailings component factors, soil condition factors, and meteorological factors. These included: Based on the correlation coefficients between the screened tailings component factors and phytoremediation strategies, the correlation coefficients between soil condition factors and phytoremediation strategies, and the correlation coefficients between meteorological factors and phytoremediation strategies, multiple individuals are generated. Specifically, when generating each individual, the tailings component factors, soil condition factors, and meteorological factors are sampled according to the absolute values ​​of the correlation coefficients between the screened tailings component factors and phytoremediation strategies, the soil condition factors, and the meteorological factors. The individual is a combination of the sampled tailings component factors, soil condition factors, and meteorological factors. Construct the fitness function; Based on the fitness function and the correlation coefficients between tailings composition factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, multiple individuals are optimized to determine the target tailings composition factors, target soil state factors, and target meteorological factors.

2. The data processing method for phytoremediation of tailings ponds according to claim 1, characterized in that, The dependent variable of the fitness function includes the correlation coefficient between the individual and the phytoremediation strategy. Specifically, for each individual, a feature vector is constructed based on the tailings composition factor, soil state factor, and meteorological factor included in the individual. The feature vector consists of the specific values ​​of the tailings composition factor, soil state factor, and meteorological factor included in the individual for the corresponding historical tailings dam phytoremediation case. For any two historical tailings dam phytoremediation cases, the similarity of the feature vectors of the individuals corresponding to the two historical tailings dam phytoremediation cases is calculated. The similarity of the feature vectors of the individuals corresponding to the individuals of any two historical tailings dam phytoremediation cases and the text similarity of the phytoremediation strategy are used as two variables and substituted into the formula for calculating the correlation coefficient to obtain the correlation coefficient between the individual and the phytoremediation strategy. The larger the absolute value of the correlation coefficient between the individual and the phytoremediation strategy, the larger the value of the fitness function.

3. The data processing method for phytoremediation of tailings ponds according to claim 2, characterized in that, Based on the correlation coefficients between fitness function and tailings composition factors and phytoremediation strategies, soil state factors and phytoremediation strategies, and meteorological factors and phytoremediation strategies, multiple individuals were optimized, including: The roulette wheel selection algorithm is used to select and retain individuals based on their fitness. Based on the correlation coefficients between tailings components, soil condition factors, and meteorological factors and phytoremediation strategies, individuals are guided to undergo crossover and mutation operations to optimize multiple individuals. Specifically, in the crossover operation, for any two selected retained individuals, the tailings component factor with the largest difference in correlation coefficient with the phytoremediation strategy, the soil condition factor with the largest difference in correlation coefficient with the phytoremediation strategy, and the meteorological factor with the largest difference in correlation coefficient with the phytoremediation strategy are exchanged. In the mutation operation, for each selected retained individual after the exchange operation, the tailings component factor, soil condition factor, and meteorological factor with the smallest absolute value of correlation coefficient in the individual are replaced with tailings component factors, soil condition factors, and meteorological factors with larger absolute values ​​of correlation coefficients that were not selected by the individual.

4. The data processing method for phytoremediation of tailings ponds according to claim 1, characterized in that, Based on the auxiliary characteristics of the tailings dam to be restored, reference cases of vegetative remediation of tailings dams were found from multiple historical tailings dam vegetative remediation cases, including: For each historical tailings dam vegetation remediation case, the remediation auxiliary characteristics of the historical tailings dam vegetation remediation case are determined based on a multidimensional remediation factor set; Calculate the similarity between the auxiliary remediation features of the tailings dam to be remediated and the auxiliary remediation features of historical tailings dam vegetation remediation cases; Based on the similarity between the auxiliary features of the tailings dam to be restored and the auxiliary features of historical tailings dam vegetation restoration cases, reference historical tailings dam vegetation restoration cases were found from multiple historical tailings dam vegetation restoration cases.

5. The data processing method for phytoremediation of tailings ponds according to claim 1, characterized in that, Collect real-time repair data for the tailings dam to be repaired, including: Multiple monitoring points were identified for the tailings dam to be repaired; Soil condition data were collected from multiple monitoring points at various remediation time points in the tailings dam to be remediated. Based on soil condition data from multiple monitoring points of the tailings dam to be repaired collected at multiple remediation time points, multiple target monitoring points were identified. Real-time repair data of the tailings dam to be repaired is collected based on multiple target monitoring points.

6. The data processing method for phytoremediation of tailings ponds according to claim 5, characterized in that, Based on soil condition data collected at multiple monitoring points in the tailings dam to be remediated at multiple remediation time points, multiple target monitoring points were identified, including: For any two monitoring points, calculate the cosine similarity and Pearson correlation coefficient of the soil state data of the tailings dam to be repaired at multiple repair time points for the two monitoring points; Based on the cosine similarity and Pearson correlation coefficient of the soil state data of the tailings pond to be repaired at multiple remediation time points of two monitoring points, multiple target monitoring points are determined. Specifically, if the cosine similarity of the soil state data of the tailings pond to be repaired at multiple remediation time points of two monitoring points is greater than the cosine similarity threshold or the Pearson correlation coefficient is greater than the Pearson correlation coefficient threshold, then one of the monitoring points is retained as the target monitoring point.

7. A data processing system for phytoremediation of tailings ponds, characterized in that, The data processing method for phytoremediation of tailings ponds according to any one of claims 1-6 includes: The case acquisition module is used to acquire multiple historical cases of phytoremediation of tailings ponds; The factor determination module is used to determine a multidimensional set of remediation factors based on multiple historical tailings dam phytoremediation cases. The multidimensional set of remediation factors includes target tailings component factors, target soil state factors, and target meteorological factors. The data acquisition module is used to collect auxiliary data for the repair of tailings dams to be repaired based on a multidimensional set of repair factors. The case search module is used to find reference historical tailings vegetation remediation cases from multiple historical tailings dam vegetation remediation cases based on the remediation auxiliary data of the tailings dam to be remediated. The strategy determination module is used to determine the phytoremediation strategy for the tailings dam to be remediated based on referenced historical tailings dam phytoremediation cases. The remediation assessment module is used to collect real-time remediation data of the tailings dam to be remediated during the execution of the phytoremediation strategy and to evaluate the remediation effect.

Images