Soil diversity recovery method constructed based on microbiota

Abstract

The invention discloses a soil diversity recovery method based on microbiota construction. The method comprises the following steps: acquiring initial microbial data and corresponding functional feature data in a preset microbial resource library based on acquired soil state data of current soil; splicing the soil state data, the initial microorganism data and the functional feature data to obtain spliced data, and obtaining a microorganism combination and a corresponding proportion of the current soil environment based on the spliced data and a trained strategy network; configuring the microorganism combination according to a corresponding proportion, and applying the configured microorganism combination to the current soil to obtain updated soil state data; and when the updated soil state data does not meet the preset remediation requirement, obtaining a new microorganism combination and a new corresponding proportion again based on the updated soil state data, the preset microorganism resource library and the trained strategy network for remediation until the preset remediation requirement is met.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of soil bioremediation, and relates to a soil diversity recovery method based on a microbial population. BACKGROUND

[0002] With the rapid development of global industrialization and urbanization, soil pollution problems are increasingly serious, posing a serious threat to ecological safety and human health. Soil pollution not only leads to soil fertility decline, crop yield reduction and quality problems, but also may cause a series of environmental problems such as groundwater pollution and loss of biodiversity.

[0003] In related technologies, physical and chemical remediation methods are used, which can improve the soil pollution condition to a certain extent, but these methods often have the disadvantages of high cost, long remediation period, secondary damage to soil structure and ecological system, etc.

[0004] Therefore, how to efficiently, stably and accurately repair the soil has become a problem to be solved. SUMMARY

[0005] Therefore, the present application provides a soil diversity recovery method based on a microbial population, which at least solves the problem that related technologies cannot efficiently, stably and accurately repair the soil.

[0006] According to a first aspect of an embodiment of the present application, a soil diversity recovery method based on a microbial population is provided, comprising: obtaining initial microbial data and corresponding functional feature data from a preset microbial resource library based on collected soil state data of a current soil; splicing the soil state data, the initial microbial data and the functional feature data to obtain spliced data, and obtaining a microbial combination and a corresponding proportion of the current soil environment based on the spliced data and a trained strategy network; configuring the microbial combination according to the corresponding proportion, and applying the configured microbial combination to the current soil to obtain updated soil state data; when the updated soil state data does not meet a preset repair requirement, obtaining a new microbial combination and a new corresponding proportion based on the updated soil state data, the preset microbial resource library and the trained strategy network for repair until the preset repair requirement is met.

[0007] According to a second aspect of the embodiments of the present application, an electronic device is provided, comprising a processor, a memory, a communication interface and a communication bus, the processor, the memory and the communication interface complete communication with each other through the communication bus; the memory is used to store at least one executable instruction, and the executable instruction causes the processor to perform the operation corresponding to the method according to the first aspect.

[0008] According to a third aspect of the embodiments of the present application, a computer storage medium is provided, and the computer storage medium stores a computer program, and the computer program is executed by a processor to implement the method according to the first aspect.

[0009] According to the scheme provided by the embodiments of the present application, the initial microorganism data and the corresponding functional characteristic data are obtained from the preset microorganism library based on the collected current soil state data of the soil; the soil state data, the initial microorganism data and the functional characteristic data are spliced to obtain spliced data, and the microorganism combination and the corresponding proportion of the current soil environment are obtained based on the spliced data and the trained strategy network; the microorganism combination is configured according to the corresponding proportion, and the configured microorganism combination is applied to the current soil to obtain updated soil state data; when the updated soil state data does not meet the preset repair requirement, a new microorganism combination and a new corresponding proportion are obtained again based on the updated soil state data, the preset microorganism library and the trained strategy network for repair until the preset repair requirement is met. In this process, the most suitable microorganism combination and proportion are dynamically selected based on the real-time soil state data and the preset microorganism library, blind application is avoided, and the repair efficiency is improved. Through the iterative feedback mechanism (detection, adjustment and reapplication), the microorganism configuration is continuously optimized, the dynamic changes in the soil repair process are adapted, and the repair effect continuously approaches the target. Combined with the functional characteristic data of the microorganism, specific problems of the soil (such as pollutant degradation and nutrient regulation) are solved, and multi-functional collaborative repair is realized. Relying on the trained strategy network, a solution is quickly generated, the cost of manual trial and error is reduced, and the feasibility of large-scale soil repair is improved. BRIEF DESCRIPTION OF DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creating any inventive labor. Figure 1 A flowchart of a soil diversity recovery method based on a microorganism population constructed according to an embodiment of the present application is provided. Figure 2A structural schematic diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0011] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application but not all the embodiments. The following embodiments are used to describe the present application but not to limit the scope of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0012] In the following description, “some embodiments” are related to a subset of all possible embodiments, but it can be understood that “some embodiments” can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict.

[0013] It should be noted that the terms “first”, “second”, “third” involved in the embodiments of the present application are only to distinguish similar objects and do not represent a specific order of the objects. Understandably, “first”, “second”, “third” can be interchanged in a specific order or sequence as allowed, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.

[0014] Those skilled in the art can understand that, unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meaning as generally understood by those skilled in the art to which the embodiments of the present application belong. It should also be understood that terms such as those defined in general dictionaries should be understood as having meanings consistent with those in the prior art, and unless specifically defined as here, should not be interpreted in an idealized or overly formal sense.

[0015] Figure 1 A flowchart of a soil diversity restoration method based on a microbial population constructed according to an embodiment of the present application is shown. The soil diversity restoration method based on a microbial population constructed according to an embodiment of the present application can be executed by an electronic device, which can be a computer, a server, etc.

[0016] As shown in Figure 1 A soil diversity restoration method based on a microbial population constructed, comprising: S101, obtaining initial microbial data and corresponding functional feature data from a preset microbial resource library based on collected soil state data of current soil.

[0017] In an embodiment of the present application, the soil state data includes soil type, soil physicochemical indicators, vegetation type, climate conditions, pollution factors; the soil type is determined by field investigation, observing the color, texture, structure and other characteristics of the soil, and combining the morphological characteristics of the soil profile. The soil is divided into sandy soil, clay soil and loamy soil. The soil physicochemical indicators are analyzed by laboratory analysis of soil pH, soil organic matter content, soil bulk density and soil oxidation-reduction potential. The vegetation type is recorded by field investigation, recording the plant species, community structure and coverage information on the soil. The climate conditions are recorded by the data provided by the weather station, recording the long-term climate characteristics and short-term climate change of the soil area. The climate conditions include temperature, precipitation, humidity, light, wind speed. The pollution factors are analyzed and detected in the laboratory, including heavy metals (such as mercury, cadmium, lead, arsenic, etc.), organic pollutants (such as petroleum pollutants, polycyclic aromatic hydrocarbons, pesticide residues, etc.), radioactive substances, pathogenic bacteria in the soil. The soil state data can be obtained by sensors, experimental analysis or other means. The preset microbial resource library is a database or resource library containing various microbial information established in advance. This resource library stores information such as the types, characteristics, adaptive environmental conditions and functions of different microorganisms. Find the initial microbial data that matches the current soil state in the preset microbial resource library. The initial microbial data can include the name, source and growth conditions of the microorganism. The functional characteristic data refers to the role of these microorganisms in the soil, such as pollutant degradation ability (such as heavy metal reduction, organic pollutant decomposition, toxicity resistance mechanism), plant growth promoting ability (nitrogen fixation, phosphorus dissolution, iron carrier production, stress resistance, salt and alkali tolerance, drought resistance, pathogenic bacteria stress resistance), pathogenic bacteria inhibition ability (antagonistic effect, antibiotic production, competitive exclusion of pathogenic bacteria), including functional tags (whether it has a certain function), metabolic pathways (involved biochemical reaction pathways), environmental adaptation parameters (reproduction speed under different culture conditions) and colonization ability (survival time in soil, competition ability).

[0018] Among them, the soil state data (such as sensor data) is received in real time through the preset platform, and when the change of the key state (such as the concentration of pollutants) exceeds the set threshold (such as 10%), the strategy model is automatically triggered to calculate and generate a response action.

[0019] S102, splice the soil state data, the initial microbial data and the functional characteristic data to obtain the spliced data, and based on the spliced data and the trained strategy network, obtain the microbial combination and the corresponding proportion of the current soil environment.

[0020] In the embodiments of the present application, the soil state data, initial microorganism data and functional feature data are spliced to obtain spliced data, the spliced data is input into the trained strategy network, and finally the microorganism combination and the corresponding proportion of the current soil environment are obtained. The corresponding proportion is the proportion of each microorganism in the combination of multiple microorganisms.

[0021] S103, configure the microorganism combination according to the corresponding proportion, and apply the configured microorganism combination to the current soil to obtain updated soil state data.

[0022] In the embodiments of the present application, the microorganism combination is mixed together according to the corresponding proportion to prepare a microorganism combination preparation. The microorganism combination can be prepared into a suspension, a powder or a granular form, etc. according to the needs, so as to facilitate the application. The microorganism combination preparation is applied to the soil by selecting a suitable application method. For example, for the preparation in the form of suspension, a spraying device can be used to uniformly spray it on the soil surface; for the powder or granular preparation, it can be applied or mixed into the soil. The application amount should be determined according to the area of the soil, the concentration of the microorganism and the expected soil restoration effect, etc. Finally, the updated soil state data is obtained.

[0023] S104, when the updated soil state data does not meet the preset repair requirement, a new microorganism combination and a new corresponding proportion are obtained again based on the updated soil state data, the preset microorganism resource library and the trained strategy network for repair until the preset repair requirement is met.

[0024] In the embodiments of the present application, the preset repair requirement can be that the concentration of pollutants in the soil is reduced to the national or local standard limit, the pH value of the soil is restored to the range suitable for plant growth, the microbial community diversity index is stable above the set threshold, and the vegetation coverage is improved to the predetermined level, etc. The updated soil state data and the preset repair requirement are compared, when the preset repair requirement is met, the repair is stopped, and subsequent real-time monitoring is performed. When the preset repair requirement is not met, the process of S101 to S103 is re-executed until the soil state data meets the preset repair requirement.

[0025] It can be understood that, in the embodiments of the present application, based on the collected current soil state data of the soil, initial microbial data and corresponding functional characteristic data are obtained from the preset microbial resource library; the soil state data, the initial microbial data and the functional characteristic data are spliced to obtain spliced data, and based on the spliced data and the trained strategy network, a microbial combination and a corresponding proportion of the current soil environment are obtained; the microbial combination is configured according to the corresponding proportion, and the configured microbial combination is applied to the current soil to obtain updated soil state data; when the updated soil state data does not meet the preset repair requirements, a new microbial combination and a new corresponding proportion are obtained again based on the updated soil state data, the preset microbial resource library and the trained strategy network for repair until the preset repair requirements are met. In this process, based on real-time soil state data and a preset microbial library, the most suitable microbial combination and proportion are dynamically selected to avoid blind application and improve repair efficiency. Through an iterative feedback mechanism (detection, adjustment and reapplication), the microbial configuration is continuously optimized to adapt to the dynamic changes in the soil repair process and ensure that the repair effect continuously approaches the target. Combined with the functional characteristic data of the microorganisms, specific problems of the soil (such as pollutant degradation and nutrient regulation) are solved to achieve multifunctional collaborative repair. Relying on the trained strategy network, a solution is quickly generated to reduce the cost of manual trial and error and improve the feasibility of large-scale soil repair.

[0026] In some embodiments of the present application, S101 can be implemented by S1011, which is described as follows.

[0027] S1011, calculate the soil state data of the current soil and the matching degree of each microorganism in the preset microbial resource library according to the attention mechanism, and obtain the initial microbial data and the corresponding functional characteristic data in the preset microbial resource library according to the matching degree.

[0028] In some embodiments of the present application, the soil state data is feature-encoded to obtain a soil feature vector, and the functional feature vector of each microorganism in the microbial resource library is obtained, and the matching degree is calculated by the attention mechanism on the soil feature vector and the functional feature vector, and the initial microbial data and the corresponding functional characteristic data are further selected from the preset microbial resource library according to the matching degree.

[0029] In some embodiments of the present application, S101 further includes S10 to S12, which are described as follows.

[0030] S10, collect a plurality of types of soil samples, and the plurality of types of soil samples include soil samples with a pollution degree greater than a preset pollution condition, clean soil samples, and soil samples with an environment harshness greater than a preset harshness condition.

[0031] In some embodiments of the present application, different types of soil samples are collected, including heavily polluted soil samples (heavy metal, petroleum hydrocarbon pollution), clean soil samples, and soil samples with harsher environmental conditions than the preset harsh conditions (saline-alkali, drought soil samples), and the sampling depth is divided into the surface layer (0-20 cm, focusing on collecting rhizosphere soil), the deep layer (20-50 cm, collecting non-rhizosphere soil), and the special depth for groundwater pollution areas (to the initial water level).

[0032] S11, obtaining a microorganism sample in the soil sample and culturing to obtain an experimental result of each microorganism sample.

[0033] S12, classifying each microorganism sample according to the test result, and constructing a preset microorganism resource library using each microorganism sample and the classification result.

[0034] In some embodiments of the present application, the obtained soil samples are cold-chain transported to the laboratory for short-term storage, and the microorganism samples in the soil samples are separated and cultured by using the culture medium to obtain the experimental results of each microorganism sample, such as whether it can degrade heavy metals, whether it can fix nitrogen, and whether it can resist drought. Each microorganism sample in the soil is classified according to the test result, and finally each microorganism sample and the classification result are stored to construct a preset microorganism resource library.

[0035] In some embodiments of the present application, S102 can be implemented by S1021 to S1023, which are described as follows.

[0036] S1021, normalizing the spliced data to obtain a normalized feature vector, and inputting the feature vector into the trained strategy network to obtain a microorganism selection probability and a recommended ratio.

[0037] S1022, based on the microorganism selection probability and a preset probability threshold, screening in the initial microorganism to obtain a candidate microorganism; and setting a corresponding ratio for the candidate microorganism through the recommended ratio.

[0038] S1023, obtaining a microorganism combination of the current soil environment based on the candidate microorganism.

[0039] In some embodiments of the present application, the microbial selection probability represents the possibility of a certain microorganism being selected in the current environment (through the Softmax output), and the recommended proportion is how much each of them should account for if multiple microorganisms are selected. After combining data of different sources and different dimensions, the normalized feature vector is obtained by uniform scaling, and the feature vector is input into the trained strategy network to obtain the microbial selection probability and the recommended proportion. The microbial selection probability is compared with the preset probability threshold, and the microorganisms below the preset probability threshold cannot be selected to obtain the candidate microorganisms. The recommended proportion is adjusted to obtain the corresponding proportion corresponding to the candidate microorganisms, and finally the microbial combination of the current soil environment is obtained based on the candidate microorganisms.

[0040] Among them, the soil state data, the initial microbial data and the functional characteristic data are spliced into a long vector in a fixed order. The soil state data has a dimension of N1=29, including soil type one-hot encoding (3 dimensions), physicochemical index standardized continuous value (4 dimensions), plant species one-hot encoding of vegetation type (top 10 dominant species, 10 dimensions) + community Shannon index (1 dimension) + coverage normalization value (1 dimension), climate condition standardized value (5 dimensions), pollution factor logarithmic conversion or normalized value (4 dimensions); the initial microbial data and the functional characteristic data have N2=39 dimensions, including functional label binary vector (20 predefined functions, 20 dimensions), metabolic path binary vector (10 dimensions), environment adaptation parameter standardized continuous value (pH, temperature, salinity range, 3 dimensions) and the like. The soil state data, the initial microbial data and the functional characteristic data are spliced into a long vector in a fixed order.

[0041] In some embodiments of the present application, S1023 can be implemented by S201, which is described by the following steps.

[0042] S201, the interaction weight between the candidate microorganisms is calculated, and when the interaction weight represents the antagonistic effect between the candidate microorganisms exceeding a preset threshold, the candidate microorganisms are reselected according to the microbial selection probability until the microbial combination and the corresponding proportion of the current soil environment are obtained.

[0043] In some embodiments of the present application, the interaction weight is whether there is a "mutual inhibition" relationship (i.e. antagonism) between microorganisms. The interaction weight between candidate microorganisms can be calculated by experimental data, literature knowledge and database preset knowledge. The interaction weight and the preset threshold are compared. When the preset threshold is exceeded, the microorganisms with a higher selection probability are selected from the remaining unselected microorganisms to replace those that will produce antagonism. After replacing the microorganisms each time, the interaction weight of the new combination is calculated again to check whether there is strong antagonism. If not, the combination can be accepted; if so, continue to adjust until a microorganism combination that neither has strong antagonism nor can exert maximum functional potential is found. Once a suitable microorganism combination is found, the recommended proportion given by the model can be used to allocate a specific use proportion to each microorganism to form the final microorganism. When constructing the microorganism resource library, the interaction relationship data (such as synergy / antagonism coefficient) between microorganisms are collected synchronously. The data can be obtained through literature mining, laboratory verification or public database, and stored in a matrix form for subsequent antagonism elimination in combination screening.

[0044] In some embodiments of the present application, during the training process of the trained strategy network, the reward function adopts a multi-objective hierarchical design, and comprehensively considers the ecological restoration effect, including pollutant removal efficiency (such as the percentage of heavy metal reduction), plant growth index improvement (such as chlorophyll content, plant height), microbial diversity improvement range (such as Shannon index change), and restoration cost (such as bacteria agent price, application difficulty). Multiple targets are quantified by weighted summation method to evaluate the influence of the current action on the soil restoration effect. The reward function The formula is: Among them, represents the restoration effect reward; represents the cost penalty; represents the change amount of pollutant removal efficiency, that is, the percentage of the concentration of target pollutants (such as heavy metals Cd, Pd) before and after restoration, which directly reflects the pollutant reduction ability of the restoration technology; represents the plant growth index improvement amount, that is, the improvement value of plant physiological or morphological index before and after restoration, including chlorophyll content change, plant height change, which is calculated by standardization and weighted summation, and is used to measure the restoration effect of soil ecological function in the restoration process (such as the ability of plant roots to fix soil and biologically enrich pollutants); represents the microbial diversity improvement amount, that is, the Shannon index change value of soil microbial community before and after restoration, which is used to measure the soil self-purification ability and anti-interference ability; Represents the repair cost value, including the price of the fungicide, the application difficulty index, and the driving algorithm reduces resource consumption under the premise of ensuring repair effect. Respectively represent the priority weight of each index, dynamically adjusted according to the repair target, regarding the dynamic weight adjustment mechanism, on the one hand, according to the repair target preset weight adjustment rule, such as setting , and refining according to the pollution type (when heavy metal repair is increased to 0.6, when organic pollution repair is kept at 0.5), and adjusting to when ecological restoration is the priority target , increase to 0.2 in cost-sensitive scenarios; on the other hand, implicitly learn the weight priority when outputting the state value through the value network, and automatically adjust the weight to maximize the reward function in training for specific repair targets, such as automatically reducing and increasing when monitoring that the plant growth index contributes less. This mechanism not only ensures quick switching of repair targets, but also realizes self-adaptation of weights in complex scenarios through data-driven, improving the response ability of the model to diversified repair needs. Through this reward function, the reinforcement learning algorithm can automatically optimize parameters such as fungicide type, application concentration, and repair period, maximizing ecological and economic benefits. The strategy network is used for strategy generation, and the state space is the input. Through two layers of fully connected network (128 neurons per layer, ReLU activation), the output is discrete microbial selection probability (Softmax) and continuous proportion parameter (Gaussian distribution), realizing the mixed action space decision of "microbial screening-proportion optimization".

[0045] Wherein, when training the strategy network and the value network, training is carried out according to the existing method, which is not described here. The training data set consists of 2000+ strains of microbial resources and corresponding functional feature data, 100+ groups of soil microcosm experimental data in laboratory controllable environment, and 300+ key parameters of soil repair related research crawled from PubMed, ScienceDirect and other databases.

[0046] Example one: For farmland soil seriously polluted by heavy metals (such as cadmium, lead), the soil diversity restoration method based on functional microbial population constructed by the present application is used for repair.

[0047] S1: Soil state data acquisition and preprocessing The collected data are as follows: 1. Soil type: through field investigation, observing soil color, texture and structure, it is judged as sandy soil.

[0048] 2. Soil physical and chemical indicators: Laboratory analysis showed that the soil pH was 6.5, the organic matter content was 1.2%, the bulk density was 1.3 g / cm 3 , and the oxidation-reduction potential was 250 mV.

[0049] 3. Vegetation type: Field surveys found that the vegetation on the soil was sparse, mainly consisting of pollution-tolerant weeds.

[0050] 4. Climate conditions: Meteorological station data showed that the annual precipitation in the area was 800 mm, and the average annual temperature was 15°C.

[0051] 5. Key pollution factors: Laboratory tests showed that the cadmium content in the soil was 5 mg / kg, exceeding the safety standard.

[0052] Data preprocessing is as follows: Categorical variables (such as soil type) are one-hot encoded, continuous variables (such as temperature, pH) are standardized by Z-score, and pollution factors (Cd available) are normalized to the [0, 1] interval.

[0053] S2: Constructing a microbial resource library Extract various functional microorganisms from the collected soil samples, such as certain sulfate-reducing bacteria and iron-reducing bacteria. Convert the extracted microorganisms into functional feature data, including functional tags (heavy metal reduction), metabolic pathways (such as sulfate reduction pathways), and environmental adaptation parameters (pH 6-8, temperature 10-35°C). Store the microorganisms and corresponding functional feature data to construct a microbial resource library, which can provide data access through corresponding interfaces.

[0054] S3: Deep reinforcement learning model outputs optimal microbial group composition and proportion Input soil state data and initial microbial data and corresponding functional feature data obtained from the microbial resource library into the DRL model. The DRL model uses the Actor-Critic framework, which means that the input soil state data and initial microbial data and corresponding functional feature data are input into the policy network (Actor) to output the optimal microbial group composition (such as sulfate-reducing bacteria A and iron-reducing bacteria B) and proportion (A:B=3:2).

[0055] S4: Field application and initial monitoring Clean the soil surface of debris, mix the microbial strains according to the output proportion of the Actor model, and prepare a suspension. Use a spraying device to evenly spray the suspension on the soil surface. Initial monitoring showed that the soil pH increased slightly, the organic matter content increased, and the cadmium content began to decrease.

[0056] S5: Data feedback and model updating When the initial monitoring data shows that the repair effect does not meet the requirements, the Actor model predicts the optimal microbial group composition and proportion in the next stage and adjusts the strategy.

[0057] Implementation effect: After several repair cycles, the cadmium content in the soil is significantly reduced, the vegetation coverage is increased, the soil microbial diversity is increased, and the soil ecological function is restored.

[0058] Reference Figure 2 The structure of the electronic device is shown in the figure, and the specific implementation of the electronic device is not limited by the embodiments of the present application.

[0059] As shown in Figure 2 The electronic device can include a processor 502, a communications interface 504, a memory 506, and a communications bus 508.

[0060] Among them: The processor 502, the communications interface 504, and the memory 506 complete mutual communication through the communications bus 508.

[0061] The communications interface 504 is configured to communicate with other electronic devices or servers.

[0062] The processor 502 is configured to execute the program 510, and specifically can execute the related steps in the above method embodiments.

[0063] Specifically, the program 510 can include program code, which includes computer operation instructions.

[0064] The processor 502 can be a central processing unit CPU, or an application specific integrated circuit ASIC, or one or more integrated circuits configured to implement embodiments of the present application. One or more processors included in the smart device can be the same type of processor, such as one or more CPUs; or can be different types of processors, such as one or more CPUs and one or more ASICs.

[0065] The memory 506 is configured to store the program 510. The memory 506 can include a high-speed RAM memory, and can also include a non-volatile memory, such as at least one disk memory.

[0066] The program 510 can specifically be used to cause the processor 502 to perform the operations corresponding to the methods described in the above method embodiments.

[0067] The specific implementation of each step in the program 510 can refer to the corresponding description in the corresponding steps and units in the method embodiments described above, and will not be described here. Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the devices and modules described above can refer to the corresponding process description in the foregoing method embodiments, and will not be described here.

[0068] It should be noted that, according to the needs of implementation, each component / step described in the embodiments of the present application can be split into more components / steps, or two or more components / steps or part of the operation of the components / steps can be combined into a new component / steps, to achieve the purpose of the embodiments of the present application.

[0069] The method according to the embodiments of the present application described above can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as CD ROM, RAM, floppy disk, hard disk or magneto-optical disk) or downloaded from a network and stored in a local recording medium, so that the method described herein can be processed by such software on a recording medium using a general-purpose computer, a special-purpose processor or programmable or special-purpose hardware (such as ASIC or FPGA). It can be understood that the computer, processor, microprocessor controller or programmable hardware includes a storage component (for example, RAM, ROM, flash memory, etc.) that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, processor or hardware, the method described herein is implemented. In addition, when the general-purpose computer accesses the code for implementing the method shown herein, the execution of the code will convert the general-purpose computer into a special-purpose computer for executing the method shown herein.

[0070] Those of ordinary skill in the art can realize that the units and method steps of each example described in connection with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether the functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of the present application.

[0071] The above embodiments are only used to illustrate the embodiments of the present application, and not to limit the embodiments of the present application. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present application, therefore all equivalent technical solutions also belong to the scope of the embodiments of the present application, the patent protection scope of the embodiments of the present application should be defined by the claims.

Claims

1. A method for soil diversity restoration based on microbial community construction, characterized in that, include: Based on the collected current soil state data, initial microbial data and corresponding functional characteristic data are obtained from the preset microbial resource library; The soil condition data, the initial microbial data, and the functional characteristic data are spliced ​​together to obtain spliced ​​data. Based on the spliced ​​data and the trained policy network, the microbial combination and corresponding proportion of the current soil environment are obtained. The microbial combination is configured according to the corresponding ratio, and the configured microbial combination is applied to the current soil to obtain updated soil state data; If the updated soil condition data does not meet the preset remediation requirements, new microbial combinations and corresponding ratios are obtained again based on the updated soil condition data, the preset microbial resource library, and the trained strategy network for remediation until the preset remediation requirements are met.

2. The method according to claim 1, characterized in that, The process of acquiring initial microbial data and corresponding functional characteristic data from a pre-set microbial resource library based on the collected current soil state data includes: The matching degree between the current soil state data and each microorganism in the preset microbial resource library is calculated based on the attention mechanism, and the initial microbial data and corresponding functional feature data are obtained from the preset microbial resource library based on the matching degree.

3. The method according to claim 1, characterized in that, Before acquiring initial microbial data and corresponding functional characteristic data from a pre-set microbial resource bank, the method based on the collected current soil state data further includes: Collect various types of soil samples, including soil samples with a pollution level greater than the preset pollution conditions, clean soil samples, and soil samples with an environmental severity greater than the preset severe conditions. Microbial samples were obtained from the soil samples and cultured to obtain experimental results for each microbial sample. The experimental results are used to classify the microbial samples by function, and the pre-set microbial resource library is constructed using the microbial samples and the classification results.

4. The method according to claim 1, characterized in that, The process of obtaining the microbial composition and corresponding proportions of the current soil environment based on the spliced ​​data and the trained policy network includes: The spliced ​​data is normalized to obtain a normalized feature vector, and the feature vector is input into the trained policy network to obtain the microbial selection probability and recommendation ratio. Based on the microbial selection probability and a preset probability threshold, candidate microorganisms are obtained by screening the initial microorganisms; and the corresponding ratio is set for the candidate microorganisms according to the recommended ratio. Based on the candidate microorganisms, obtain the microbial composition of the current soil environment.

5. The method according to claim 4, characterized in that, The process of obtaining the microbial composition of the current soil environment based on the candidate microorganisms includes: The interaction weights between the candidate microorganisms are calculated. When the interaction weights, which represent the antagonistic effects between the candidate microorganisms, exceed a preset threshold, candidate microorganisms are reselected based on the microorganism selection probability until the current microbial combination and corresponding proportion of the soil environment are obtained.

6. The method according to claim 1, characterized in that, The reward function of the trained policy network is set as follows during the training process: r In the above formula, These represent the priority weights of each indicator. This indicates the change in pollutant removal efficiency. Indicates the improvement in plant growth indicators. This indicates an increase in microbial diversity. This indicates the cost and value of repair.

Images