Development of data processing methods based on artificial intelligence technology

Abstract

The present invention discloses an artificial intelligence-based technology development data processing method, which belongs to the field of data processing technology. The method obtains technology development data processing tasks input by technical personnel, and then uses the verified technology development data processing tasks as a basis to perform task learning and analysis using artificial intelligence algorithms to determine technology development data processing rules. Finally, real-time technology development data can be processed using the technology development data processing rules. This method can not only effectively improve the processing efficiency of technology development data, but also effectively reduce the error-prone problems caused by human processing of technology development data.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and specifically relates to a technology development data processing method based on artificial intelligence. Background Art

[0002] With the rapid development of information technology, the era of big data has arrived. Extracting valuable information from this massive amount of data has become a pressing issue. Technology development data refers to the various types of data generated during the research and development process. This data is crucial for technological advancement, product innovation, and process optimization. Existing technologies often rely on staff to analyze technology development data independently or wait for a problem to occur before analyzing it. This manual processing of technology development data is inefficient, inaccurate, and prone to errors. Summary of the Invention

[0003] The present invention provides a technology development data processing method based on artificial intelligence to solve the problems existing in the prior art.

[0004] A technology development data processing method based on artificial intelligence, comprising:

[0005] Acquire a technology development data processing task input by a technician, and verify the technology development data processing task to obtain a verified technology development data processing task;

[0006] Based on the verified technology development data processing tasks, artificial intelligence algorithms are used to conduct task learning analysis and determine the technology development data processing rules;

[0007] Real-time technology development data is collected, and the real-time technology development data is processed according to technology development data processing rules to obtain processed technology development data.

[0008] Furthermore, obtaining the technology development data processing task input by the technical personnel includes: obtaining the batch technology development data input by the technical personnel and the data label corresponding to each technology development data, and obtaining the technology development data processing task.

[0009] Furthermore, the technology development data processing task is verified to obtain the verified technology development data processing task, including:

[0010] For each type of technology development data in the technology development data processing task, verify whether it is numerical data. If so, perform data length verification; otherwise, convert the text data into numerical data and then perform data length verification;

[0011] For each type of technical development data in the technical development data processing task, check whether the length is the same. If so, obtain the verified technical development data processing task; otherwise, perform length consistency processing to obtain the verified technical development data processing task.

[0012] Furthermore, based on the verified technology development data processing tasks, artificial intelligence algorithms are used to conduct task learning analysis to determine technology development data processing rules, including:

[0013] A uniform initialization strategy is used to initialize the hyperparameters of the artificial intelligence algorithm and obtain multiple hyperparameter codes corresponding to the artificial intelligence algorithm;

[0014] Based on the verified technical development data processing tasks, the fitness corresponding to each hyperparameter encoding is obtained, and multiple hyperparameter encodings are divided into optimal encoding, elite encoding, general encoding, and inferior encoding according to the fitness of each hyperparameter encoding;

[0015] For the optimal coding, a global mutation optimization strategy is used to perform a global greedy search on the optimal coding, and the optimal coding after the global greedy search is obtained;

[0016] For elite codes, an adaptive fluctuation optimization strategy is used to perform random local search on the elite codes to obtain the elite codes after random local search.

[0017] For general codes, a multi-code collaborative optimization strategy is used to conduct joint local search on the general codes to obtain the general codes after the joint local search;

[0018] For poor quality codes, an adaptive sinusoidal fast guidance strategy is used to conduct a fast guidance search for the poor quality codes, and the poor quality codes after the fast guidance search are obtained;

[0019] The best code after global greedy search, the elite code after random local search, the general code after joint local search, and the inferior code after fast guided search are re-fused into a population;

[0020] Determine whether the current number of learning times is greater than or equal to the maximum number of learning times. If so, obtain the final hyperparameters corresponding to the artificial intelligence algorithm based on the reintegrated population, and use the artificial intelligence algorithm after applying the final hyperparameters as the technical development data processing rules. Otherwise, based on the reintegrated population, return to the hyperparameter coding and division steps.

[0021] Furthermore, a uniform initialization strategy is used to initialize the hyperparameters of the artificial intelligence algorithm to obtain multiple hyperparameter codes corresponding to the artificial intelligence algorithm, including:

[0022] For the hyperparameters of the artificial intelligence algorithm, they are randomly initialized within the upper and lower limits of the hyperparameters, and the initialized hyperparameters are encoded into vectors to obtain basic encodings;

[0023] Using the basic code as the first hyperparameter code, N hyperparameter codes are obtained: ;in, Indicates the i Hyperparameter encoding, represents the first constant coefficient, and Set to 0.5; represents the second constant coefficient, and Set to 0.2; represents pi, represents the remainder function, represents the sine function, Indicates the i +1 hyperparameter encoding, i =1,2,…,M-1.

[0024] Furthermore, based on the verified technical development data processing task, the fitness corresponding to each hyperparameter encoding is obtained, and multiple hyperparameter encodings are divided into optimal encoding, elite encoding, general encoding and inferior encoding according to the fitness of each hyperparameter encoding, including:

[0025] Using the verified technical development data in the technical development data processing task as input data and the data label corresponding to the verified technical development data in the technical development data processing task as the expected output, obtaining the error function value corresponding to the hyperparameter encoding;

[0026] Take the negative value of the error function corresponding to the hyperparameter encoding to obtain the fitness corresponding to the hyperparameter encoding;

[0027] The hyperparameter encoding with the largest fitness is taken as the optimal encoding, and the remaining hyperparameter encodings are arranged in descending order of fitness to obtain a hyperparameter encoding sequence;

[0028] The hyperparameter encoding sequence is divided into elite encoding, general encoding and inferior encoding in sequence using a ratio of 2:6:2.

[0029] Furthermore, for the optimal coding, a global mutation optimization strategy is used to perform a global greedy search on the optimal coding, and the optimal coding after the global greedy search is obtained, including:

[0030] Based on the current number of learning times, the global mutation probability is obtained as: ;in, represents the global mutation probability, represents the second-type Euler integral function, represents pi, represents the mutation probability control constant, Indicates the current number of learning times;

[0031] For the optimal code, based on the global mutation probability, a roulette wheel method is used to determine the mutation operation selection result corresponding to the optimal code; wherein the mutation operation selection result includes whether mutation is required or not;

[0032] When the mutation operation selection result is that no mutation is required, the original optimal code is directly used as the optimal code after the global greedy search;

[0033] When the mutation operation selection result indicates that mutation is required, the optimal encoding is mutated as follows: ;in, Indicates the t The optimal encoding in the learning process, Represents the optimal encoding after mutation , Represents the first random number between (0,1);

[0034] Determine whether the fitness of the optimal code after mutation increases. If so, use the optimal code after mutation as the optimal code after global greedy search. Otherwise, use the original optimal code as the optimal code after global greedy search.

[0035] Furthermore, for the elite code, an adaptive fluctuation optimization strategy is used to perform a random local search on the elite code, and the elite code after the random local search is obtained, including:

[0036] Based on the current number of learning times, the hyperbolic tangent function is used to obtain the adaptive fluctuation factor: ;in, represents the adaptive fluctuation factor, represents the cosine function, represents the hyperbolic tangent function, Indicates the preset maximum number of learning times. represents pi;

[0037] A random local search is performed on the elite code according to the adaptive fluctuation factor, and the elite code after the random local search is obtained as follows: ;in, Indicates the t Second learning process j Elite code, j =1,2,…,N1, N1 represents the total number of elite codes, represents the second random number between (0,1), represents the elite encoding after random local search , e represents a natural constant, sin represents a sine function, and cos represents a cosine function.

[0038] Furthermore, for general codes, a multi-code collaborative optimization strategy is used to perform a joint local search on the general codes, and the general codes after the joint local search are obtained, including:

[0039] For any general code, randomly match an elite code to the general code to obtain the target elite code corresponding to the general code;

[0040] Obtain the Euclidean distance between the general code and its corresponding target elite code, and obtain the random movement factor based on the Euclidean distance: ;in, Expressed as the general code d The random shift factor generated by the dimension parameter, Indicates the preset maximum number of learning times. Represents a random number between (-1,1), Indicates the m The Euclidean distance between a general code and its corresponding target elite code, d =1,2,…,L, L represents the total dimension of the parameter;

[0041] According to the optimal code, random movement factor and target elite code, a joint local search is performed on the general code, and the general code after the joint local search is obtained as follows: ;in, Indicates the t The first m The general code d dimensional parameters, Indicates the m The first general code after the joint local search d dimensional parameters, represents the first learning factor, represents the second learning factor, represents the third random number between (0,1), represents the fifth random number between (0,1), Indicates the t The first m The first target elite code corresponding to the general code d dimensional parameters, m =1,2,…,N2, N2 represents the total number of general codes.

[0042] Furthermore, for the poor quality codes, an adaptive sinusoidal fast guiding strategy is used to perform a fast guiding search on the poor quality codes, and the poor quality codes after the fast guiding search are obtained, including:

[0043] Generate adaptive inertia weight and adaptive step search factor as: as well as ;in, represents the adaptive inertia weight, Expressed as a natural constant e The exponential function with base , Indicates the preset maximum number of learning times. represents the adaptive step search factor, represents a constant factor;

[0044] According to the adaptive inertia weight and the adaptive step size search factor, a fast guided search is performed on the inferior code, and the inferior code after the fast guided search is obtained as follows: ;in, Indicates the t The first n A poor quality code, Indicates poor quality encoding after a fast boot search , means (0,2 ), represents pi, Represents the seventh random number between (0,1).

[0045] The present invention provides an artificial intelligence-based technology development data processing method, which obtains technology development data processing tasks input by technical personnel, and then uses artificial intelligence algorithms to perform task learning and analysis based on the verified technology development data processing tasks to determine technology development data processing rules. Finally, real-time technology development data can be processed according to the technology development data processing rules. This can not only effectively improve the processing efficiency of technology development data, but also effectively reduce the error-prone problems caused by manual processing of technology development data. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0047] Figure 1 A flowchart of a data processing method based on artificial intelligence technology development provided by an embodiment of the present invention.

[0048] Figure 2 A flowchart of determining technical development data processing rules provided by an embodiment of the present invention.

[0049] The above drawings illustrate specific embodiments of the present invention, which will be described in more detail below. These drawings and the accompanying description are not intended to limit the scope of the present invention in any way, but rather to illustrate the concept of the present invention to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0050] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0051] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0052] like Figure 1 As shown, an embodiment of the present invention provides a method for processing technology development data based on artificial intelligence, including:

[0053] S11. Obtaining a technology development data processing task input by a technician, and verifying the technology development data processing task to obtain a verified technology development data processing task;

[0054] For example, if the technical development data is the parameters for designing a motor, when the operating temperature or speed needs to be simulated based on the motor parameters, modeling is usually required, which requires technical developers to have high-level professional knowledge. However, artificial intelligence technology can achieve rapid prediction, allowing staff to adjust the parameters to a more reasonable value before conducting detailed testing, which can significantly improve technical development efficiency. Alternatively, if the technical development data is the test data for a software (which may include input parameters, software control parameters, output parameters, etc.), and then the technical development data needs to be classified, artificial intelligence technology can be used to quickly classify the test data, thereby improving the storage and search efficiency of the technical development data.

[0055] S12. Based on the verified technology development data processing tasks, use artificial intelligence algorithms to conduct task learning analysis and determine technology development data processing rules;

[0056] By learning the relationship between data and labels in technology development data processing tasks, we can effectively assist technical personnel in quickly classifying technology development data, thereby improving the processing efficiency of technology development data.

[0057] S13: Collect real-time technology development data, and process the real-time technology development data according to technology development data processing rules to obtain processed technology development data.

[0058] For example, it is necessary to develop a load forecasting software for a power consumption area. The technical development data can be the factors affecting power consumption, and these factors need to be processed. At this time, the corresponding expected output is the actual power consumption corresponding to the power consumption factors. After learning these data, load forecasting and technical development can be achieved.

[0059] The present invention provides an artificial intelligence-based technology development data processing method, which obtains technology development data processing tasks input by technical personnel, and then uses artificial intelligence algorithms to perform task learning and analysis based on the verified technology development data processing tasks to determine technology development data processing rules. Finally, real-time technology development data can be processed according to the technology development data processing rules. This can not only effectively improve the processing efficiency of technology development data, but also effectively reduce the error-prone problems caused by manual processing of technology development data.

[0060] In an embodiment of the present invention, obtaining a technical development data processing task input by a technician includes: obtaining a batch of technical development data input by the technician and a data label corresponding to each technical development data, to obtain the technical development data processing task.

[0061] For example, a development data may include development data B1, development data B2 and development data B3, and the corresponding label is C. Then the technical development data processing task may be: input technical development data [B1, B2, B3], and the expected output data should be C.

[0062] In an embodiment of the present invention, verifying the technology development data processing task to obtain the verified technology development data processing task includes:

[0063] For each type of technology development data in the technology development data processing task, verify whether it is numerical data. If so, perform data length verification; otherwise, convert the text data into numerical data and then perform data length verification;

[0064] For each type of technical development data in the technical development data processing task, check whether the length is the same. If so, obtain the technical development data processing task after verification. Otherwise, perform length consistency processing (such as filling zeros in front of the data) to obtain the technical development data processing task after verification.

[0065] like Figure 2As shown, based on the verified technology development data processing task, artificial intelligence algorithms are used to perform task learning analysis and determine technology development data processing rules, including:

[0066] S21. Initialize the hyperparameters of the artificial intelligence algorithm using a uniform initialization strategy to obtain multiple hyperparameter codes corresponding to the artificial intelligence algorithm;

[0067] S22. Based on the verified technical development data processing task, obtain the fitness corresponding to each hyperparameter encoding, and classify the multiple hyperparameter encodings into optimal encoding, elite encoding, general encoding, and inferior encoding according to the fitness of each hyperparameter encoding;

[0068] S23, for the optimal code, using a global mutation optimization strategy to perform a global greedy search on the optimal code, and obtaining the optimal code after the global greedy search;

[0069] S24, for the elite code, using an adaptive fluctuation optimization strategy to perform a random local search on the elite code, and obtaining an elite code after the random local search;

[0070] S25. For the general code, a multi-code collaborative optimization strategy is used to perform a joint local search on the general code to obtain the general code after the joint local search;

[0071] S26. For the low-quality code, use an adaptive sinusoidal fast guidance strategy to perform a fast guidance search on the low-quality code to obtain the low-quality code after the fast guidance search;

[0072] S27, re-merge the optimal code after the global greedy search, the elite code after the random local search, the general code after the joint local search, and the inferior code after the fast guided search into a population;

[0073] S28. Determine whether the current number of learning times is greater than or equal to the maximum number of learning times. If so, obtain the final hyperparameters corresponding to the artificial intelligence algorithm based on the reintegrated population, and use the artificial intelligence algorithm after applying the final hyperparameters as the technical development data processing rules. Otherwise, return to the step of hyperparameter coding and division based on the reintegrated population.

[0074] When artificial intelligence algorithms are currently used for data classification or data prediction, gradient descent algorithms are often used to optimize the hyperparameters corresponding to the artificial intelligence algorithms. This often leads to falling into local optimality, resulting in poor data classification or data prediction effects. For technical development data, if the classification is wrong, it is very likely to cause errors in the subsequent test optimization process. Therefore, an embodiment of the present invention provides an improved hyperparameter optimization algorithm to improve the accuracy of technical development data.

[0075] In an embodiment of the present invention, a uniform initialization strategy is used to initialize the hyperparameters of the artificial intelligence algorithm to obtain multiple hyperparameter codes corresponding to the artificial intelligence algorithm, including:

[0076] For the hyperparameters of the artificial intelligence algorithm, they are randomly initialized within the upper and lower limits of the hyperparameters, and the initialized hyperparameters are encoded into vectors to obtain basic encodings;

[0077] Using the basic code as the first hyperparameter code, N hyperparameter codes are obtained: ;in, Indicates the i Hyperparameter encoding, represents the first constant coefficient, and Set to 0.5; represents the second constant coefficient, and Set to 0.2; represents pi, represents the remainder function, represents the sine function, Indicates the i +1 hyperparameter encoding, i =1,2,…,M-1.

[0078] The uniform initialization strategy provided by the embodiment of the present invention can make the hyperparameter encoding more evenly distributed in the initial solution space, thereby effectively improving the optimization speed of the algorithm.

[0079] In the embodiment of the present invention, based on the verified technical development data processing task, the fitness corresponding to each hyperparameter encoding is obtained, and multiple hyperparameter encodings are divided into optimal encoding, elite encoding, general encoding and inferior encoding according to the fitness of each hyperparameter encoding, including:

[0080] Using the verified technical development data in the technical development data processing task as input data and the data label corresponding to the verified technical development data in the technical development data processing task as the expected output, obtaining the error function value corresponding to the hyperparameter encoding;

[0081] Take the negative value of the error function corresponding to the hyperparameter encoding to obtain the fitness corresponding to the hyperparameter encoding;

[0082] The hyperparameter encoding with the largest fitness is taken as the optimal encoding, and the remaining hyperparameter encodings are arranged in descending order of fitness to obtain a hyperparameter encoding sequence;

[0083] The hyperparameter code sequence is divided into elite code, general code, and poor quality code in a ratio of 2:6:2. To ensure integer partitioning, the number of hyperparameter codes should be an integer multiple of 10 plus one.

[0084] In an embodiment of the present invention, a global mutation optimization strategy is used to perform a global greedy search on the optimal encoding, and the optimal encoding after the global greedy search is obtained, including:

[0085] Based on the current number of learning times, the global mutation probability is obtained as: ;in, represents the global mutation probability, represents the second-type Euler integral function, represents pi, represents the mutation probability control constant, Indicates the current number of learning times;

[0086] For the optimal code, based on the global mutation probability, a roulette wheel method is used to determine the mutation operation selection result corresponding to the optimal code; wherein the mutation operation selection result includes whether mutation is required or not;

[0087] When the mutation operation selection result is that no mutation is required, the original optimal code is directly used as the optimal code after the global greedy search;

[0088] When the mutation operation selection result indicates that mutation is required, the optimal encoding is mutated as follows: ;in, Indicates the t The optimal encoding in the learning process, Represents the optimal encoding after mutation , Represents the first random number between (0,1);

[0089] Determine whether the fitness of the optimal code after mutation increases. If so, use the optimal code after mutation as the optimal code after global greedy search. Otherwise, use the original optimal code as the optimal code after global greedy search.

[0090] The technical development data provided by the embodiment of the present invention can provide a larger mutation probability in the early and middle stages of the algorithm, and the rate of decrease is slower. After the algorithm reaches the late stage, the mutation probability decreases rapidly, thereby improving the convergence accuracy of the algorithm. At the same time, combined with greedy search, the convergence speed of the algorithm is further guaranteed.

[0091] In an embodiment of the present invention, an adaptive fluctuation optimization strategy is used to perform a random local search on the elite code to obtain the elite code after the random local search, including:

[0092] Based on the current number of learning times, the hyperbolic tangent function is used to obtain the adaptive fluctuation factor: ;in, represents the adaptive fluctuation factor, represents the cosine function, represents the hyperbolic tangent function, Indicates the preset maximum number of learning times. represents pi;

[0093] A random local search is performed on the elite code according to the adaptive fluctuation factor, and the elite code after the random local search is obtained as follows: ;in, Indicates the t Second learning process j Elite code, j =1,2,…,N1, N1 represents the total number of elite codes, represents the second random number between (0,1), represents the elite encoding after random local search , e represents a natural constant, sin represents a sine function, and cos represents a cosine function.

[0094] The adaptive fluctuation optimization strategy provided by the embodiment of the present invention can enable the elite code to search in different directions and at different speeds in the surrounding areas, which can effectively search for better areas and improve the possibility of the algorithm searching for the optimal solution.

[0095] In an embodiment of the present invention, a multi-code collaborative optimization strategy is used to perform a joint local search on a general code, and the general code obtained after the joint local search is obtained, including:

[0096] For any general code, randomly match an elite code to the general code to obtain the target elite code corresponding to the general code;

[0097] Obtain the Euclidean distance between the general code and its corresponding target elite code, and obtain the random movement factor based on the Euclidean distance: ;in, Expressed as the general code d The random shift factor generated by the dimensional parameter, Indicates the preset maximum number of learning times. Represents a random number between (-1,1), Indicates the m The Euclidean distance between a general code and its corresponding target elite code, d =1,2,…,L, where L represents the total dimension of the parameter;

[0098] According to the optimal code, random movement factor and target elite code, a joint local search is performed on the general code, and the general code after the joint local search is obtained as follows: ;in, Indicates the t The first m The first generally coded d dimensional parameters, Indicates the m The first general code after the joint local search d dimensional parameters, represents the first learning factor, represents the second learning factor, represents the third random number between (0,1), represents the fifth random number between (0,1), Indicates the t The first m The first target elite code corresponding to the general code d dimensional parameters, m =1,2,…,N2, N2 represents the total number of general codes.

[0099] The multi-coding collaborative optimization strategy provided by the embodiment of the present invention can effectively fuse the coding information and, with a certain degree of disturbance, can more effectively search for unfamiliar areas between coding individuals, improve the traversal of the solution space, and enhance the algorithm search capability.

[0100] In an embodiment of the present invention, for poor-quality codes, an adaptive sinusoidal fast-steering strategy is used to perform a fast-steering search on the poor-quality codes, and the poor-quality codes after the fast-steering search are obtained, including:

[0101] Generate adaptive inertia weight and adaptive step search factor as: as well as ;in, represents the adaptive inertia weight, Expressed as a natural constant e The exponential function with base , Indicates the preset maximum number of learning times. represents the adaptive step search factor, represents a constant factor;

[0102] According to the adaptive inertia weight and the adaptive step size search factor, a fast guided search is performed on the inferior code, and the inferior code after the fast guided search is obtained as follows: ;in, Indicates the t The first n A poor quality code, Indicates poor quality encoding after a fast boot search , means (0,2 ), represents pi, Represents the seventh random number between (0,1).

[0103] The adaptive sinusoidal fast guidance strategy provided by the embodiment of the present invention uses a sinusoidal waveform to quickly advance toward the optimal area, which can not only effectively improve the search capability of the algorithm, but also improve the global search capability.

[0104] Optionally, after each update of the code, the code may be processed for out-of-bounds errors to ensure parameter validity.

[0105] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. It should be understood that the present invention is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and variations can be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A technology development data processing method based on artificial intelligence, characterized in that: include: Acquire a technology development data processing task input by a technician, and verify the technology development data processing task to obtain a verified technology development data processing task; Based on the verified technology development data processing tasks, artificial intelligence algorithms are used to conduct task learning analysis and determine the technology development data processing rules; collecting real-time technology development data, and processing the real-time technology development data using technology development data processing rules to obtain processed technology development data; Based on the verified technology development data processing tasks, artificial intelligence algorithms are used to conduct task learning analysis and determine technology development data processing rules, including: A uniform initialization strategy is used to initialize the hyperparameters of the artificial intelligence algorithm and obtain multiple hyperparameter codes corresponding to the artificial intelligence algorithm; Based on the verified technical development data processing tasks, the fitness corresponding to each hyperparameter encoding is obtained, and multiple hyperparameter encodings are divided into optimal encoding, elite encoding, general encoding, and inferior encoding according to the fitness of each hyperparameter encoding; For the optimal coding, a global mutation optimization strategy is used to perform a global greedy search on the optimal coding, and the optimal coding after the global greedy search is obtained; For elite codes, an adaptive fluctuation optimization strategy is used to perform random local search on the elite codes to obtain the elite codes after random local search. For general codes, a multi-code collaborative optimization strategy is used to conduct joint local search on the general codes to obtain the general codes after the joint local search; For poor quality codes, an adaptive sinusoidal fast guidance strategy is used to conduct a fast guidance search for the poor quality codes, and the poor quality codes after the fast guidance search are obtained; The best code after global greedy search, the elite code after random local search, the general code after joint local search, and the inferior code after fast guided search are re-fused into a population; Determine whether the current number of learning times is greater than or equal to the maximum number of learning times. If so, obtain the final hyperparameters corresponding to the artificial intelligence algorithm based on the reintegrated population, and use the artificial intelligence algorithm after applying the final hyperparameters as the technical development data processing rules. Otherwise, based on the reintegrated population, return to the hyperparameter coding and division steps.

2. The method for processing technology development data based on artificial intelligence according to claim 1, characterized in that: The task of processing the technology development data input by the technical personnel is obtained, including: obtaining the batch technology development data input by the technical personnel and the data label corresponding to each technology development data, to obtain the technology development data processing task.

3. The method for processing technology development data based on artificial intelligence according to claim 1, characterized in that: Verifying the technology development data processing task to obtain the verified technology development data processing task includes: For each type of technology development data in the technology development data processing task, verify whether it is numerical data. If so, perform data length verification; otherwise, convert the text data into numerical data and then perform data length verification; For each type of technical development data in the technical development data processing task, check whether the length is the same. If so, obtain the verified technical development data processing task; otherwise, perform length consistency processing to obtain the verified technical development data processing task.

4. The method for processing technology development data based on artificial intelligence according to claim 1, characterized in that: The uniform initialization strategy is used to initialize the hyperparameters of the artificial intelligence algorithm and obtain multiple hyperparameter codes corresponding to the artificial intelligence algorithm, including: For the hyperparameters of the artificial intelligence algorithm, they are randomly initialized within the upper and lower limits of the hyperparameters, and the initialized hyperparameters are encoded into vectors to obtain basic encodings; Using the basic code as the first hyperparameter code, N hyperparameter codes are obtained: ;in, Indicates the i Hyperparameter encoding, represents the first constant coefficient, and Set to 0.5; represents the second constant coefficient, and Set to 0.2; represents pi, represents the remainder function, represents the sine function, Indicates the i +1 hyperparameter encoding, i =1,2,…,M-1.

5. The method for processing technology development data based on artificial intelligence according to claim 1, characterized in that: Based on the verified technical development data processing tasks, the fitness corresponding to each hyperparameter encoding is obtained, and multiple hyperparameter encodings are divided into optimal encoding, elite encoding, general encoding and poor quality encoding according to the fitness of each hyperparameter encoding, including: Using the verified technical development data in the technical development data processing task as input data and the data label corresponding to the verified technical development data in the technical development data processing task as the expected output, obtaining the error function value corresponding to the hyperparameter encoding; Take the negative value of the error function corresponding to the hyperparameter encoding to obtain the fitness corresponding to the hyperparameter encoding; The hyperparameter encoding with the largest fitness is taken as the optimal encoding, and the remaining hyperparameter encodings are arranged in descending order of fitness to obtain a hyperparameter encoding sequence; The hyperparameter encoding sequence is divided into elite encoding, general encoding and inferior encoding in sequence using a ratio of 2:6:

2.

6. The method for processing technology development data based on artificial intelligence according to claim 1, characterized in that: For the optimal coding, a global mutation optimization strategy is used to perform a global greedy search on the optimal coding, and the optimal coding after the global greedy search is obtained, including: Based on the current number of learning times, the global mutation probability is obtained as: ;in, represents the global mutation probability, represents the second-type Euler integral function, represents pi, represents the mutation probability control constant, Indicates the current number of learning times; For the optimal code, based on the global mutation probability, a roulette wheel method is used to determine the mutation operation selection result corresponding to the optimal code; wherein the mutation operation selection result includes whether mutation is required or not; When the mutation operation selection result is that no mutation is required, the original optimal code is directly used as the optimal code after the global greedy search; When the mutation operation selection result indicates that mutation is required, the optimal encoding is mutated as follows: ;in, Indicates the t The optimal encoding in the learning process, Represents the optimal encoding after mutation , Represents the first random number between (0,1); Determine whether the fitness of the optimal code after mutation increases. If so, use the optimal code after mutation as the optimal code after global greedy search. Otherwise, use the original optimal code as the optimal code after global greedy search.

7. The method for processing technology development data based on artificial intelligence according to claim 6, characterized in that: For elite codes, an adaptive fluctuation optimization strategy is used to perform random local search on the elite codes, and the elite codes after random local search are obtained, including: Based on the current number of learning times, the hyperbolic tangent function is used to obtain the adaptive fluctuation factor: ;in, represents the adaptive fluctuation factor, represents the cosine function, represents the hyperbolic tangent function, Indicates the preset maximum number of learning times. represents pi; A random local search is performed on the elite code according to the adaptive fluctuation factor, and the elite code after the random local search is obtained as follows: ;in, Indicates the t Second learning process j Elite code, j =1,2,…,N1, N1 represents the total number of elite codes, represents the second random number between (0,1), represents the elite encoding after random local search , e represents a natural constant, sin represents a sine function, and cos represents a cosine function.

8. The method for processing technology development data based on artificial intelligence according to claim 7, characterized in that: For general codes, a multi-code collaborative optimization strategy is used to conduct a joint local search on the general codes, and the general codes obtained after the joint local search include: For any general code, randomly match an elite code to the general code to obtain the target elite code corresponding to the general code; Obtain the Euclidean distance between the general code and its corresponding target elite code, and obtain the random movement factor based on the Euclidean distance: ;in, Expressed as the general code d The random shift factor generated by the dimension parameter, Indicates the preset maximum number of learning times. Represents a random number between (-1,1), Indicates the m The Euclidean distance between a general code and its corresponding target elite code, d =1,2,…,L, where L represents the total dimension of the parameter; According to the optimal code, random movement factor and target elite code, a joint local search is performed on the general code, and the general code after the joint local search is obtained as follows: ;in, Indicates the t The first m The general code d dimensional parameters, Indicates the m The first general code after the joint local search d dimensional parameters, represents the first learning factor, represents the second learning factor, represents the third random number between (0,1), represents the fifth random number between (0,1), Indicates the t The first m The first target elite code corresponding to the general code d dimensional parameters, m =1,2,…,N2, N2 represents the total number of general codes.

9. The method for processing technology development data based on artificial intelligence according to claim 8, characterized in that: For poor-quality codes, an adaptive sinusoidal fast-guiding strategy is used to quickly guide the search for poor-quality codes. The poor-quality codes obtained after the fast-guiding search include: Generate adaptive inertia weight and adaptive step search factor as: as well as ;in, represents the adaptive inertia weight, Expressed as a natural constant e The exponential function with base , Indicates the preset maximum number of learning times. represents the adaptive step search factor, represents a constant factor; According to the adaptive inertia weight and the adaptive step size search factor, a fast guided search is performed on the inferior code, and the inferior code after the fast guided search is obtained as: ;in, Indicates the t The first n A poor quality code, Indicates poor quality encoding after a fast boot search , means (0,2 ), represents pi, Represents the seventh random number between (0,1).

Images