A test design method and system for affecting the production of bacillus subtilis spores

Abstract

The present disclosure provides a test design method and system for affecting the production of bacillus subtilis spores, relating to the technical field of statistical test design, comprising obtaining basic test information of a test, determining a relationship model between a test response and influencing factors; according to the value range of each factor and the number of each factor level, the candidate set of each factor and the design space of the test are determined; using a recursive search strategy, starting from all support points in the current test design, recursively searching for points with larger gradients in the test space, and adding these points to the test design to obtain the next second test design; adjusting the weight of the design points in the second test design to obtain the third test design, using the optimal function to check whether the third test design meets the optimal criteria, if it meets, the current test design is the optimal test design. The problem of low quality of feasible solutions under the condition of limited calculation time is solved.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of statistical experiment design, in particular to a test design method and system for affecting spore yield of Bacillus subtilis. BACKGROUND

[0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute the prior art.

[0003] The response concerned in the test of the influence of the culture condition on the spore yield of Bacillus subtilis is the spore yield of Bacillus subtilis, and the test design is used to prove the influence of the culture condition on the spore yield of Bacillus subtilis. The test design is a mathematical principle and implementation method for formulating a proper test scheme according to a predetermined goal to facilitate effective statistical analysis of test results. In many industrial fields and scientific research fields, there is a need to conduct large-scale multivariate and multilevel tests, and since the test requires time and money costs, there is a need to simulate the entire test area with as few test times as possible. If the test scheme is well designed, the test times and costs can be reduced, and representative test results can be obtained; if the test scheme is not well designed, increasing the test times may not result in better results.

[0004] At present, there are various test design methods, such as optimal test design, orthogonal test design, Latin square test design, factor design, uniform test design, etc. Test design has been widely used in industrial applications and computer simulation. Among them, the optimal test design can save test costs and further reduce test times compared with other test design methods.

[0005] Regarding the optimal test design, various construction methods have been proposed by different scholars. Mainly based on mathematical principles, such as the Fedorov-Wynn method, the Multiplicative method, and this year, a faster Randomized Exchange method has been proposed by a scholar, but like the Fedorov-Wynn method and the Multiplicative method, these methods have the problem of too long convergence time, especially for constructing optimal test design for multivariate and multilevel tests, which consumes a lot of time, and the quality of feasible solutions cannot be guaranteed under limited computing time. SUMMARY

[0006] In order to solve the above problems, the present disclosure proposes a test design method and system for affecting the spore yield of Bacillus subtilis, which solves the problem of slow convergence speed of the traditional optimal test design construction method and low quality of feasible solutions under limited computing time.

[0007] According to some embodiments, the present disclosure adopts the technical solutions as follows:

[0008] A test design method for affecting the spore yield of Bacillus subtilis, comprising:

[0009] According to the purpose, conditions and basic test information obtained from experience of the test for affecting the spore yield of Bacillus subtilis, a relationship model between the test response and the influencing factors is determined.

[0010] According to the value range and the number of levels of each factor, a candidate set of each factor and a design space of the test are determined.

[0011] A plurality of support points are randomly selected in the test space to form an initial test design, and then a recursive search strategy is used to recursively search for points with larger gradients in the test space from all support points in the current test design, and these points are added to the test design to obtain a second test design of the next step. The design points in the second test design are adjusted in weight to obtain a third test design, and an optimal function is used to check whether the third test design meets the optimal criteria, and if it does, the current test design is the optimal test design.

[0012] The optimal test design is converted into a test scheme for testing the influence on the spore yield of Bacillus subtilis.

[0013] According to some embodiments, the present disclosure adopts the technical solutions as follows:

[0014] A test design system for affecting the spore yield of Bacillus subtilis, comprising:

[0015] An initialization model construction module is configured to determine a relationship model between a test response and influencing factors according to the purpose, conditions and basic test information obtained from experience of the test for affecting the spore yield of Bacillus subtilis.

[0016] A test design space construction module is configured to determine a candidate set of each factor and a design space of the test according to the value range and the number of levels of each factor.

[0017] An optimal recursive search module is configured to randomly select a plurality of support points in the test space to form an initial test design, and then use a recursive search strategy to recursively search for points with larger gradients in the test space from all support points in the current test design, and add these points to the test design to obtain a second test design of the next step. The design points in the second test design are adjusted in weight to obtain a third test design, and an optimal function is used to check whether the third test design meets the optimal criteria, and if it does, the current test design is the optimal test design. The optimal test design is converted into a test scheme for testing the influence on the spore yield of Bacillus subtilis.

[0018] According to some embodiments, the present disclosure adopts the technical solutions as follows:

[0019] A non-transitory computer readable storage medium for storing computer instructions, which, when executed by a processor, implement the test design method for affecting spore yield of Bacillus subtilis.

[0020] According to some embodiments, the present disclosure adopts the technical solutions as follows:

[0021] An electronic device, comprising a processor, a memory and a computer program; wherein the processor is connected with the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the test design method for affecting spore yield of Bacillus subtilis.

[0022] Compared with the prior art, the present disclosure has the beneficial effects that:

[0023] The present disclosure provides an optimal test design construction method based on recursive search in the test design method for affecting spore yield of Bacillus subtilis, which solves the problem of slow convergence speed of the traditional optimal test design construction method and low quality of feasible solutions under the condition of limited calculation time.

[0024] At the same time, the present method has low sensitivity to the number of test levels, and can construct optimal test design for multi-factor and multi-level test. DETAILED DESCRIPTION

[0025] The accompanying drawings, which form a part of the present disclosure, are used to provide a further understanding of the present disclosure, and the illustrative embodiments thereof and their descriptions serve to explain the present disclosure, and do not constitute improper limitations on the present disclosure.

[0026] Figure 1 The flowchart of the optimal test design construction method based on recursive search of the present disclosure embodiment;

[0027] Figure 2 The process schematic diagram of creating 2-factor test neighborhood space of the present disclosure embodiment;

[0028] Figure 3 The process schematic diagram of creating 3-factor test neighborhood space of the present disclosure embodiment;

[0029] Figure 4 The D-optimal test design equivalence theorem diagram of the present disclosure for affecting spore yield of Bacillus subtilis under culture conditions. DETAILED DESCRIPTION

[0030] The present disclosure will be further described with reference to the drawings and examples.

[0031] It should be noted that the following detailed description is exemplary in nature and is intended to provide further description of the present disclosure. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure belongs.

[0032] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments in accordance with the present disclosure. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, steps, operations, elements, components, and / or groups thereof, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof.

[0033] Example 1

[0034] In one embodiment of the present disclosure, a test design method for affecting the yield of Bacillus subtilis spores is provided, comprising:

[0035] Step 1: According to the purpose, conditions and experience of the test for affecting the yield of Bacillus subtilis spores, determine the relationship model between the test response and the influencing factors;

[0036] Step 2: According to the value range and the number of levels of each factor, determine the candidate set of each factor and the design space of the test;

[0037] Step 3: Randomly select a plurality of support points in the test space to form an initial test design, then use a recursive search strategy to recursively search for points with larger gradients in the test space from all support points in the current test design, and add these points to the test design to obtain a second test design. Adjust the weights of the design points in the second test design to obtain a third test design, and use an optimal function to check whether the third test design meets the optimal criteria. If it does, the current test design is the optimal test design.

[0038] As an embodiment, the basic test information includes the test response, the influencing factors, the value range of each factor, the number of levels of each factor, and the constraint relationship between each factor.

[0039] As an embodiment, a D-optimal test design is constructed for the test for affecting the yield of Bacillus subtilis spores under culture conditions, and the optimal test design construction method of the recursive search comprises the following steps:

[0040] Step 1: Based on the experimental objective, conditions, and engineering experience, provide basic experimental information, including the experimental response (i.e., the quality characteristic of interest), influencing factors and their value ranges, the number of levels for each factor, the constraints between factors, and the relationship model between the response and the factors; the specific method for determining the relationship model between the response and the factors is as follows:

[0041] Based on the influencing factors of the experiment and the basic information of the response experiment, a mathematical model is used to present the relationship model between the response and the factors.

[0042] The relational model is denoted as:

[0043] y = f1(x)β1 + f2(x)β2 + ... + f m (x)β m +ε

[0044] Where f1(x), f2(x), ..., f m (x) is a continuous function in the design space, ε~N(0,σ). 2 The error is represented by y = β. The relational model can be simplified as y = β. T f(x) + ε. Where y is the response; x = (x1, x2, ..., x...) n ) represents an n-dimensional vector composed of various factors, where n is the number of factors in the experiment; β = (β1, β2, ..., β m ) T Let f(x) be a vector consisting of m parameters to be estimated; f(x) = (f1(x), f2(x), ..., f m (x)) T The parameter values ​​are guesses given based on past engineering experience, reflecting the influence relationship between various factors and the response.

[0045] Step 2: Based on the value range of each factor and the number of levels for each factor, use a grid. i Let i = 1, 2, ..., n, and determine the candidate set χ for each factor. i And determine the design space χ for the experiment.

[0046] The design space of the experiment is the combination of all possible levels of influencing factors, i.e., χ = χ1 × χ2 × … × χ n For all possible levels of factors, containing One point.

[0047] Step 3: Randomly select multiple support points in the experimental space to form the initial experimental design; specifically, randomly select multiple support points in the experimental space, set the weight of each support point to form the initial experimental design, and calculate the information matrix of the initial experimental design. The experimental design can be represented as a set composed of support points and weights, and the sum of the weights of all support points is 1.

[0048] As an embodiment, in step 3, m support points are randomly selected in the trial space χ, each support point has a weight of 1 / m, to form an initial design ξ0, and the information matrix is calculated, and the information matrix calculation formula is N represents the number of support points in the trial design. In addition, a non-negative integer s is used to represent the trial design index, and the trial design ξ s Can be represented as a set composed of support points x j and weights w j (j=1,2,…,N), and the sum of the weights of all support points is 1;

[0049] Step 4: Using a recursive strategy, recursively search for points with larger gradients in the trial space from all support points in the current trial design, and add these points to the trial design to obtain the next second trial design. The steps are as follows:

[0050] Set the empty set collection and set the support point as the current neighborhood center, starting from the current neighborhood center, construct a neighborhood search space, find the point with the largest gradient in the neighborhood search space, and compare its gradient with the gradient of the current neighborhood center;

[0051] If the gradient of the point with the largest gradient in the search space is greater than the gradient of the current neighborhood center, set the point with the largest gradient in the search space as the current neighborhood center.

[0052] As an embodiment, using a recursive search strategy, recursively search for points with larger gradients in the trial space from all support points in the current trial design ξ s , and add these points to the trial design to obtain the next second trial design ξ s+1 .

[0053] Step 4.1: Set the set T to be empty;

[0054] From all support points in the current design ξ s , execute step 4.2, step 4.3, step 4.4 and step 4.5 in turn;

[0055] Step 4.2: Set the support point x j as the current neighborhood center x center ;

[0056] Step 4.3: From the current neighborhood center x center , construct a search space χ search .

[0057]

[0058] Where, indicates the neighborhood center xcenter The kth factor.

[0059] like Figure 2 As shown, in an experiment with two factors, each factor having a value range of [0,10] and 11 levels per factor, the process of constructing the neighborhood search space starting from the experimental point (5,5) as the neighborhood center is as follows: x center =(5,5),

[0060] χ search ={(0,5),(1,5),(2,5),(3,5),(4,5),(6,5),(7,5),(8,5),(9,5),(10,5),(5,0),(5,1),(5,2),(5,3),(5,4),(5,6),(5,7),(5,8),(5,9),(5,10)}.

[0061] like Figure 3 As shown, in an experiment with 3 factors, each with a value range of [0,10] and 11 levels, the process of constructing the neighborhood search space starting from the experimental point (5,5,5) as the neighborhood center is as follows: x center =(5,5,5),

[0062] χ search ={(5,5,0),(5,5,1),(5,5,2),(5,5,3),(5,5,4),(5,5,6),(5,5,7),(5,5,8),(5,5,9),(5,5,10),5,0,5),(5,1,5),(5,2,5),(5,3,5),(5,4,5),(5,6,5),(5,7,5),(5,8,5),(5,9,5),(5,10,5),(0,5,5),(1,5,5),(2,5,5),(3,5,5),(4,5,5),(6,5,5),(7,5,5),(8,5,5),(9,5,5),(10,5,5)}.

[0063] Step 4.4: In the search space χ search Find the point with the largest gradient in And compare its gradient with the gradient of the current neighborhood center. If The gradient is greater than x center The gradient will then Set the current neighborhood center as the current center, and recursively execute steps 4.3 and 4.4; where the gradient formula for the design point is:

[0064]

[0065] The gradient function g(x) can reflect the improvement of the test design in the solution space after moving in a direction.

[0066] Step 4.5: If is not a support point of the current design ξ s , then is inserted into the set T.

[0067] Step 4.6: All points in the set T are added to the current test design to generate the next second test design ξ s+1 .

[0068] This step follows the formula:

[0069] ξ s+1 = (1-α)ξ s + αξ T

[0070] where T size and are the size of the set T and the size of ξ s , respectively, and ξ T is the design composed of all points in the set T, and the weight of all support points in ξ T is

[0071] Step 5: Adjust the weights of the design points in the current second test design to obtain the next third test design. The adjustment process follows the formula:

[0072]

[0073] where is the weight of support point j in the test design generated in the s-th iteration.

[0074] Step 6: Use the optimal function φ(x, ξ) to check whether the current test design meets the optimal criteria. If it does, the current test design is the optimal test design. If it does not, repeat steps 4, 5, and 6. The optimal function is:

[0075]

[0076] where g(x) - φ(x, ξ) is the gradient target value. When the gradient value of the test design ξ is less than or equal to 0 for all test points, the test design is the optimal test design. That is, when the value of the optimal function at all test points is less than or equal to 0, and all points equal to 0 are support points of the test design, the test design is the optimal test design.

[0077] When combined as the D-optimal test design method for the test of the influence of culture conditions on the spore yield of Bacillus subtilis, the specific process method is as follows:

[0078] Step Q1: The response concerned in the test of the influence of culture conditions on the spore yield of Bacillus subtilis is the spore yield of Bacillus subtilis, and the test factors include the methanol content, the ethanol content, the propanol content, the butanol content, the pH value and the culture time in the culture conditions. The value range of each factor is shown in Table 1, and the level number of each test factor is 11. Based on the test basic information such as the test factors and the response, the relationship model between the response and the factors is given in the form of a mathematical model, y = β0+ β1x1+ β2x2+ β3x3+ β4x4+ β5x5+ β6x6+ β7x2x3+ ε, β = (β1, β2, β3, β4, β5, β6, β7). The parameter guess is given according to the past engineering experience, β = (27.79, 9.0, 4.27, 1.0, 1.8, -3.07, 4.63, -1.77).

[0079] Table 1 Value range of each test factor in the test of the influence of culture conditions on the spore yield of Bacillus subtilis

[0080]

[0081] Step Q2: According to the value range of each factor and the level number of each factor, the candidate set χ1, χ2, χ3, χ4, χ5, χ6 of each factor and the test space χ are determined.

[0082] χ1 = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}

[0083] χ2 = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}

[0084] χ3 = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}

[0085] χ4 = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10}

[0086] χ5 = {6.0, 6.3, 6.6, 6.9, 7.2, 7.5, 7.8, 8.1, 8.4, 8.7, 9.0}

[0087] χ6 = {1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2.0}.

[0088] Step Q3: 7 test points are randomly selected in the test space χ to form the initial test design ξ0.

[0089]

[0090] Then, in step Q4: using a recursive search strategy, sequentially search from the current experimental design ξ... s Starting from all support points, recursively search for points with larger gradients in the experimental space, and add these points to the experimental design to obtain the next experimental design ξ. s+1 .

[0091] Step Q4.1: Set set T to an empty set;

[0092] Starting sequentially from the current design ξ s Starting from all support points, execute steps Q4.2, Q4.3, Q4.4, and Q4.5;

[0093] Step Q4.2: Set support point x j For the current neighborhood center x center ;

[0094] Step Q4.3: From the current neighborhood center x center Starting from this point, construct the neighborhood search space. in, Represents the neighborhood center x center The kth factor.

[0095] In an experiment with two factors, each with values ​​ranging from [0,10] and 11 levels, the process of constructing the neighborhood search space starting from the experimental point (5,5) as the neighborhood center is as follows: Figure 2 As shown.

[0096] In an experiment with 3 factors, each with values ​​ranging from [0,10] and 11 levels, the process of constructing the neighborhood search space starting from the experimental point (5,5) as the neighborhood center is as follows: Figure 3 As shown.

[0097] Step Q4.4: In the search space χ search Find the point with the largest gradient in And compare its gradient with the gradient of the current neighborhood center. If The gradient is greater than x center The gradient will then Set the current neighborhood center as the current center, and recursively execute steps Qq4.3 and Q4.4; where the gradient formula for the design point is:

[0098]

[0099] Step Q4.5: If Not the current design ξ s The support point will be Inserting all points in set T into the current design;

[0100] Step Q4.6: Inserting all points in set T into the current design to generate the next design ξ s+1 . This step follows the formula ξ s+1 = (1 - a)ξ s + aξ T , where T size and are the size of set T and the size of ξ s , respectively, and ξ T is the design consisting of all points in set T, with all support points in ξ T having weight

[0101] Step Q5: Adjusting the weights of the design points in the current design ξ s+1 to obtain the next design ξ s+2 , following the formula:

[0102]

[0103] Step Q6: Checking the current design using the optimal function φ(x, ξ) to see if it meets the optimal criteria. If it does, the current design is the optimal design. If it does not, repeat steps Q4, Q5, and Q6. The optimal function is:

[0104]

[0105] A design ξ is optimal when the value of the optimal function φ(x, ξ) is less than or equal to zero at all design points, and all points where the value is zero are support points of the design ξ.

[0106] The D-optimal design obtained by the above method is shown in Table 2.

[0107] Table 2 D-optimal design table for the experiment on the effect of culture conditions on the spore yield of Bacillus subtilis

[0108]

[0109]

[0110]

[0111] Compared with the randomized exchange method, the present method can construct an optimal design in a shorter time, as shown in Table 2. The Fedorov-Wynn method and the multiplication method cannot construct an optimal design within an acceptable time, so they are not compared.

[0112] Table 3: Comparison of time consumption and optimal efficiency of the present method and randomized exchange method

[0113] Method Time (sec) Optimal efficiency The method 110.08 0.9999990033303406 Randomized exchange method 719.13 0.9999999667245463

[0114] wherein the optimal efficiency is calculated by the formula:

[0115]

[0116] wherein, x max = argmax g(x). The ratio of the optimal efficiency test design ξ and the statistical efficiency of the optimal test design.

[0117] Embodiment 2

[0118] In an embodiment of the present disclosure, a test design system for affecting the yield of Bacillus subtilis spores is provided, comprising:

[0119] An initialization model construction module is configured to determine a relationship model between a test response and an influencing factor according to a purpose, a condition and basic test information of the test for affecting the yield of Bacillus subtilis spores;

[0120] A test design space construction module is configured to determine a candidate set of each factor and a design space of the test according to a value range and a level number of each factor;

[0121] An optimal recursive search module is configured to randomly select a plurality of support points in the test space to form an initial test design, then use a recursive search strategy to sequentially start from all support points in the current test design to recursively search for points with larger gradients in the test space, and add these points to the test design to obtain a second test design of the next step; adjust the weight of the design points in the second test design to obtain a third test design, and use an optimal function to check whether the third test design meets an optimal criterion, and if so, the current test design is the optimal test design.

[0122] Embodiment 3

[0123] In an embodiment of the present disclosure, a non-transitory computer readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the test design method for affecting the yield of Bacillus subtilis spores.

[0124] Embodiment 4

[0125] An electronic device, comprising: a processor, a memory and a computer program; wherein the processor is connected with the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the test design method for affecting the yield of Bacillus subtilis spores.

[0126] The present disclosure is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to embodiments of the present disclosure. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the computer or other programmable data processing devices generate a device that realizes the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one flow or multiple flows and / or blocks Figure 1 The functions specified in one flow or multiple flows and / or blocks

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device to cause a series of operation steps to be performed on the computer or other programmable data processing device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable data processing device provide a process for realizing the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one flow or multiple flows and / or blocks Figure 1 The functions specified in one flow or multiple flows and / or blocks

[0128] Although the specific embodiments of the present disclosure are described above with reference to the accompanying drawings, the present disclosure is not limited to the above embodiments, and various modifications or changes can be made to the embodiments without departing from the scope of the present disclosure.

Claims

1. A method of designing an experiment affecting the production of Bacillus subtilis spores, characterized in that, The application comprises the following steps: According to the purpose, conditions and basic test information of the Bacillus subtilis spore yield influence test, the relationship model between the test response and the influence factors is determined; According to the value range of each factor and the number of each factor level, the candidate set of each factor and the design space of the test are determined; A plurality of support points are randomly selected in the test space to form an initial test design, and then a recursive search strategy is used to recursively search for points with larger gradients in the test space from all support points in the current test design, and these points are added to the test design to obtain a second test design in the next step; The design points in the second test design are adjusted in weight to obtain a third test design, and an optimal function is used to check whether the third test design meets the optimal criteria, and if it meets the optimal criteria, the current test design is the optimal test design; The optimal test design is converted into a test scheme for testing the influence on the yield of Bacillus subtilis spores; The step of using the recursive strategy to recursively search for points with larger gradients in the test space from all support points in the current test design, and adding these points to the test design to obtain a second test design in the next step comprises the following steps: An empty set is set, and a support point is set as a current neighborhood center, and a neighborhood search space is constructed from the current neighborhood center to find a point with the largest gradient in the neighborhood search space, and the gradient of the point is compared with the gradient of the current neighborhood center; If the gradient of the point with the largest gradient in the search space is greater than the gradient of the current neighborhood center, the point with the largest gradient in the search space is set as the current neighborhood center.

2. A design of experiments method for affecting the production of Bacillus subtilis spores as claimed in claim 1, wherein, The basic test information comprises a test response, influence factors, a value range of each factor, a number of levels of each factor, and a constraint relationship between each factor.

3. A design of experiments method for affecting the production of Bacillus subtilis spores as claimed in claim 1, wherein, The method for determining the relationship model between the test response and the influence factors comprises: based on the basic information of the influence factors and the response test, a mathematical model is used to give the relationship model between the response and the factors.

4. A design of experiments method for affecting the production of Bacillus subtilis spores as claimed in claim 1, wherein, The design space of the test is all possible combinations of the levels of the influence factors.

5. A design of experiments method for affecting the production of Bacillus subtilis spores as claimed in claim 1, wherein, If the point with the largest gradient in the search space is not a support point of the current test design, the point with the largest gradient in the search space is inserted into the empty set, all points in the empty set are added to the current test design to generate a second test design in the next step.

6. A design of experiments method for affecting the production of Bacillus subtilis spores as claimed in claim 1, wherein, The application comprises the following steps:

7. A test design system for affecting production of Bacillus subtilis spores, characterized by, The initialization model construction module is used to determine the relationship model between the test response and the influence factors according to the purpose, conditions and basic test information of the Bacillus subtilis spore yield influence test; The test design space construction module is used to determine the candidate set of each factor and the design space of the test according to the value range of each factor and the number of each factor level; ​ The optimal recursive search module is configured to randomly select a plurality of support points in the trial space to form an initial trial design, and then use a recursive search strategy to sequentially start from all support points in the current trial design to recursively search for points with larger gradients in the trial space, and add these points to the trial design to obtain a second trial design for the next step; The weight of the design points in the second trial design is adjusted to obtain a third trial design, and an optimal function is used to check whether the third trial design meets an optimal criterion, and if so, the current trial design is the optimal trial design; and the optimal trial design is converted into a test scheme for testing the influence of Bacillus subtilis spore yield. The step of using the recursive strategy to sequentially start from all support points in the current trial design to recursively search for points with larger gradients in the trial space, and add these points to the trial design to obtain a second trial design for the next step comprises the following steps: Set an empty set and set the support point as the current neighborhood center, start from the current neighborhood center to construct a neighborhood search space, find the point with the largest gradient in the neighborhood search space, and compare the gradient of the point with the gradient of the current neighborhood center; If the gradient of the point with the largest gradient in the search space is greater than the gradient of the current neighborhood center, set the point with the largest gradient in the search space as the current neighborhood center.

8. A non-transitory computer-readable storage medium, comprising: The non-transitory computer readable storage medium is configured to store computer instructions, which are executed by a processor to implement the trial design method for testing the influence of Bacillus subtilis spore yield according to any one of claims 1-6.

9. An electronic device, comprising: It comprises: A processor, a memory and a computer program; wherein the processor is connected with the memory, and the computer program is stored in the memory; when the electronic device is running, the processor executes the computer program stored in the memory, so that the electronic device executes the trial design method for testing the influence of Bacillus subtilis spore yield according to any one of claims 1-6.

Images