Aerospace engineering data algorithm model full life cycle management method and system

Abstract

The invention provides an aerospace engineering data algorithm model full life cycle management method and system, and relates to the technical field of aerospace engineering algorithm model management and service. According to the method provided by the invention, a plurality of different types of algorithm model files are stored in a model library, and an inference flow chart corresponding to a target data analysis flow is generated in response to an editing operation and a configuration operation input by a user; and in response to a release operation input by a user, generating and sending an algorithm model service file corresponding to the target data analysis process to a server according to the reasoning flow chart, so that the server deploys an algorithm model service according to the algorithm model service file. According to the method provided by the invention, the deployment of the corresponding algorithm model service can be quickly completed based on the analysis configuration operation and the algorithm model file set by the user, and then the analysis processing of the data is completed through the algorithm model service, so that the model service construction cost is reduced, the model application efficiency is improved, and the data analysis processing capability is enhanced.

Description

Technical Field

[0001] The present invention relates to the field of aerospace engineering algorithm model management and service technology, and in particular to a full life cycle management method and system for aerospace engineering data algorithm model. Background Art

[0002] In the field of manned space engineering, with the long-term on-orbit operation of space stations and the implementation of subsequent missions, massive engineering data is generated. Data mining and analysis research is carried out based on these data, and a large number of algorithm model files for data analysis and processing are obtained. When technicians need to analyze and process data by using algorithm model files, due to the complex processes of algorithm model file management, deployment, monitoring and evaluation, the model management platform in related technologies has the following problems: decentralized heterogeneous model management, lack of a unified coordination mechanism; complex model deployment process, long time cycle; imperfect model monitoring mechanism, difficult risk control, etc. As a result, technicians cannot quickly deploy corresponding algorithm model services through algorithm model files, data analysis and processing requires a lot of time, and there are problems such as low data processing efficiency and poor model application effect.

[0003] Therefore, there is an urgent need for a full life cycle management method and system for aerospace engineering data algorithm models, which can quickly complete the deployment and update of corresponding algorithm model services based on algorithm model files, thereby reducing the cost of model service construction, improving the efficiency of model application, and enhancing data analysis and processing capabilities. Summary of the invention

[0004] The embodiments of the present invention provide a method and system for the full life cycle management of aerospace engineering data algorithm models, which can quickly complete the deployment and update of corresponding algorithm model services based on algorithm model files, thereby reducing the cost required for model service construction, improving the efficiency of model application, and enhancing data analysis and processing capabilities.

[0005] To achieve the above object, the embodiments of the present invention adopt the following technical solutions: In a first aspect, a method for managing the entire life cycle of an aerospace engineering data algorithm model is provided, which is applied to a first device, and the first device communicates with a server. The method includes: storing a plurality of algorithm model files of different categories in a model library; generating an inference flowchart corresponding to a target data analysis process in response to editing operations and configuration operations input by a user in a preset interface; the inference flowchart includes a plurality of nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logical nodes; the editing operation includes adding a node operation, deleting a node operation, and / or connecting a node operation; the configuration operation includes determining the algorithm model file corresponding to each model node from the model library and / or configuring each logical node; The execution code corresponding to the point, the model node is used to process the input data according to the configured algorithm model file, and the logic node is used to process the input data according to the configured execution code; in response to the publishing operation input by the user in the preset interface, the algorithm model service file corresponding to the target data analysis process is generated according to the inference flowchart and sent to the server, so that the server deploys the algorithm model service according to the algorithm model service file, the algorithm model service is used to respond to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, processes the data to be analyzed according to the target data analysis process, obtains the data analysis result, and sends the data analysis result to the second device.

[0006] In a possible implementation of the first aspect, the logic node includes a data processing node, a routing node and a combiner node; the data processing node includes at least one input terminal and at least one output terminal, the data processing node is used to process the data input by the input terminal according to the configured execution code, and output the data analysis result through the output terminal; the combiner node includes at least one input terminal and multiple output terminals, the combiner node is used to output the data input by the input terminal through multiple output terminals respectively according to the configured execution code; the routing node includes at least one input terminal and multiple output terminals, the routing node is used to determine the target output terminal according to the data input by the input terminal through the configured execution code, and output the input data through the target output terminal, and the target output terminal is one of the multiple output terminals.

[0007] In a possible implementation of the first aspect, after the server deploys the algorithm model service according to the algorithm model service file, the method further includes: sending an operation status monitoring instruction to the server, the operation status monitoring instruction is used to instruct: the server obtains the operation status parameters corresponding to the algorithm model service, the operation status parameters include memory utilization, processor utilization and response time; the server generates a first alarm message when any operation status parameter is greater than or equal to a first threshold and the duration is greater than or equal to a second threshold, the first alarm message is used to prompt that the operation status of the algorithm model service is abnormal; the server sends the first alarm message to the first device at a preset time interval.

[0008] In a possible implementation of the first aspect, after the server deploys the algorithm model service according to the algorithm model service file, the method further includes: obtaining verification data, the verification data including multiple verification samples and a first analysis result obtained by each verification sample based on the target data analysis process; sending an analysis result evaluation instruction to the server, the analysis result evaluation instruction carrying the verification data, the analysis result evaluation instruction being used to instruct: the server processes the multiple verification samples according to the target data analysis process based on the algorithm model service to obtain a second analysis result corresponding to each verification sample; the server determines an evaluation indicator corresponding to the algorithm model service based on the first analysis result and the second analysis result corresponding to each verification sample, and generates a second report when any evaluation indicator of the server is greater than or equal to a preset threshold. The server sends a second alarm message to the first device, and the second alarm message is used to prompt that the analysis result of the algorithm model service is abnormal; where, when the category of the algorithm model service is a binary classification model, the evaluation indicators include: accuracy, precision, recall, specificity, area under the ROC curve, KS test value and F1 measure; where the category of the algorithm model service is a multi-classification model, the evaluation indicators include: accuracy, weighted precision, weighted recall and weighted F value; where the category of the algorithm model service is a regression model, the evaluation indicators include: mean absolute error, mean square error, root mean square error and determination coefficient; where the category of the algorithm model service is a clustering model, the evaluation indicators include: silhouette coefficient-cosine similarity, silhouette coefficient-Euclidean distance, residual sum of squares and covariance.

[0009] In a possible implementation of the first aspect, the method further includes: in response to an iterative update instruction input by a user, iteratively training the algorithm model files corresponding to one or more model nodes included in the algorithm model service according to the training code to obtain a trained algorithm model file; updating the algorithm model service file according to the trained algorithm model file to obtain an updated algorithm model service file; and sending the updated algorithm model service file to the server, so that the server deploys the algorithm model service according to the updated algorithm model service file.

[0010] In a possible implementation of the first aspect, the method further includes: sending a data distribution difference monitoring instruction to the server, the data distribution difference monitoring instruction carrying training data for each algorithm model file included in the algorithm model service, the data distribution difference monitoring instruction being used to instruct: upon receiving a data analysis request sent by the second device, the server determines the data distribution difference value between the data to be analyzed and the training data carried in the data analysis request, and when the data distribution difference value is greater than or equal to a third threshold, the server generates a third alarm message, and the server sends the third alarm message to the second device, the third alarm message being used to prompt that there is a data distribution difference between the data to be analyzed and the training data.

[0011] In a possible implementation of the first aspect, the algorithm model service is also used to respond to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, adds the data to be analyzed to a task queue arranged in order, processes the multiple data to be analyzed included in the task queue in order according to the target data analysis process, obtains the data analysis result corresponding to each data to be analyzed, and sends the data analysis result corresponding to each data to be analyzed to the second device.

[0012] In a possible implementation of the first aspect, in a case where a server is deployed with multiple algorithm model services corresponding to a target data analysis process, the method further includes: sending a publishing mode setting instruction to the server, the publishing mode setting instruction is used to instruct: the server sets the publishing mode of the multiple algorithm model services corresponding to the target data analysis process; wherein the publishing modes include independent publishing, grayscale publishing or comparative publishing; in a case where the publishing mode is independent publishing, the publishing mode setting instruction carries an identifier of the target algorithm model service, the target algorithm model service is one of the multiple algorithm model services, the server responds to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, the server processes the data to be analyzed through the target algorithm model service, obtains a data analysis result, and sends the data analysis result to the second device; When the publishing mode is grayscale publishing, the server responds to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, the server processes the data to be analyzed through multiple algorithm model services to obtain the data analysis results corresponding to each algorithm model service, and sends the data analysis results corresponding to each algorithm model service to the second device; when the publishing mode is comparative publishing, the server responds to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, the server processes the data to be analyzed through multiple algorithm model services to obtain the data analysis results corresponding to each algorithm model service, the server determines the evaluation index corresponding to each algorithm model service, and sends the data analysis result of the algorithm model service corresponding to the maximum value of the evaluation index corresponding to each algorithm model service to the second device.

[0013] The beneficial effects of the present invention are as follows: the method provided by the present invention stores multiple different types of algorithm model files in a model library; generates a reasoning flowchart corresponding to the target data analysis process in response to the editing operation and configuration operation input by the user in the preset interface; generates and sends the algorithm model service file corresponding to the target data analysis process to the server according to the reasoning flowchart in response to the publishing operation input by the user in the preset interface, so that the server deploys the algorithm model service according to the algorithm model service file. The method provided by the present invention can quickly complete the deployment of the corresponding algorithm model service based on the analysis configuration operation and the algorithm model file set by the user, and then complete the analysis and processing of the data through the algorithm model service, reduce the cost required for the construction of the model service, improve the efficiency of the model application, and enhance the data analysis and processing capabilities. It can also be understood that by uniformly managing a variety of different types of heterogeneous model files, the centralized management and maintenance of the algorithm model files are realized, and the management efficiency is improved. In response to the operation input by the user, the reasoning logic diagram corresponding to the target data analysis process is configured, and the algorithm model service corresponding to the target data analysis process is quickly deployed, which can effectively simplify the deployment process of the service and shorten the cycle from development to deployment of the algorithm model service. In addition, the method provided by the present invention can monitor the running status of the algorithm model service in real time, and ensure the stability and reliability of the model during operation. And the processing results of the algorithm model service are evaluated through multi-dimensional evaluation indicators, providing a comprehensive analysis of the performance of the algorithm model service and accurately grasping the performance of the algorithm model service. According to the monitoring and evaluation results, the algorithm model service is iteratively updated through a visualization page, which speeds up the iteration speed of the model and further enhances the performance and adaptability of the model. The method provided by the present invention effectively controls the risks of the model in the production environment through the management of the entire life cycle of the model, improves the comprehensive management level of the model, and ensures the continuity and convenience of the model application.

[0014] In a second aspect, the present invention provides a full life cycle management system for aerospace engineering data algorithm models, which is applied to a first device, and the first device communicates with a server. The system includes: a model management module, which is used to store multiple algorithm model files of different categories in a model library; a service configuration module, which is used to generate an inference flowchart corresponding to a target data analysis process in response to editing operations and configuration operations input by a user in a preset interface; the inference flowchart includes multiple nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logical nodes; the editing operation includes adding a node operation, deleting a node operation and / or connecting a node operation; the configuration operation includes determining the algorithm model file corresponding to each model node from the model library and / or Configure the execution code corresponding to each logical node. The model node is used to process the input data according to the configured algorithm model file. The logical node is used to process the input data according to the configured execution code. The service publishing module is used to respond to the publishing operation input by the user in the preset interface, generate and send the algorithm model service file corresponding to the target data analysis process to the server according to the inference flowchart, so that the server deploys the algorithm model service according to the algorithm model service file. The algorithm model service is used to respond to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, processes the data to be analyzed according to the target data analysis process, obtains the data analysis result, and sends the data analysis result to the second device.

[0015] According to a third aspect, an electronic device is provided, comprising a memory and one or more processors; the memory is coupled to the processor; wherein the memory stores computer program code, the computer program code comprises computer instructions, and when the computer instructions are executed by the processor, the electronic device executes a method as in any implementation of the first aspect.

[0016] According to a fourth aspect, a computer-readable storage medium is provided, comprising computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes the method in any implementation of the first aspect.

[0017] According to a fifth aspect, a computer program product is provided. When the computer program product is run on a computer, the computer is enabled to execute the method in any implementation of the first aspect.

[0018] It can be understood that the beneficial effects that can be achieved by the system of the second aspect, the electronic device of the third aspect, the computer-readable storage medium of the fourth aspect, and the computer program product of the fifth aspect provided above can be referred to the beneficial effects in the first aspect and any possible design method thereof, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1A schematic diagram of a system architecture provided by an embodiment of the present invention; Figure 2 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention; Figure 3 A flowchart of a method for managing the entire life cycle of an aerospace engineering data algorithm model provided by an embodiment of the present invention; Figure 4 A schematic diagram of a preset interface provided by an embodiment of the present invention; Figure 5 A schematic diagram of an inference flow chart provided by an embodiment of the present invention; Figure 6 A flowchart of a method for managing the entire life cycle of an aerospace engineering data algorithm model provided by an embodiment of the present invention; Figure 7 A flowchart of a method for managing the entire life cycle of an aerospace engineering data algorithm model provided by an embodiment of the present invention; Figure 8 A schematic diagram of a node parameter configuration interface provided by an embodiment of the present invention; Fig. 9 A flowchart of a method for managing the entire life cycle of an aerospace engineering data algorithm model provided by an embodiment of the present invention; Fig.10 A schematic diagram of the structure of a management system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The technical solution in the embodiment of the present invention will be described below in conjunction with the accompanying drawings in the embodiment of the present invention. In the description of the present invention, unless otherwise specified, " / " indicates that the objects associated before and after are in an "or" relationship. For example, A / B can represent A or B; the "or" in the present invention is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A or B can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. In addition, in the description of the present invention, unless otherwise specified, "multiple" means two or more than two. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items.

[0021] In addition, in order to clearly describe the technical solutions of the embodiments of the present invention, in the embodiments of the present invention, words such as "first" and "second" are used to distinguish the same or similar items with substantially the same functions and effects. Those skilled in the art can understand that words such as "first" and "second" do not limit the quantity and execution order, and words such as "first" and "second" do not necessarily limit the difference.

[0022] Meanwhile, in the embodiments of the present invention, words such as "exemplary" or "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present invention should not be interpreted as being better or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a concrete way for easy understanding.

[0023] In the field of manned space engineering, with the long-term on-orbit operation of space stations and the implementation of subsequent missions, massive engineering data is generated. Data mining and analysis research is carried out based on these data, and a large number of algorithm model files for data analysis and processing are obtained. When technicians need to analyze and process data by using algorithm model files, due to the complex processes of algorithm model file management, deployment, monitoring and evaluation, the model management platform in related technologies has the following problems: decentralized heterogeneous model management, lack of a unified coordination mechanism; complex model deployment process, long time cycle; imperfect model monitoring mechanism, difficult risk control, etc. As a result, technicians cannot quickly deploy corresponding algorithm model services through algorithm model files, data analysis and processing requires a lot of time, and there are problems such as low data processing efficiency and poor model application effect.

[0024] Therefore, there is an urgent need for a full life cycle management method and system for aerospace engineering data algorithm models, which can quickly complete the deployment and update of corresponding algorithm model services based on algorithm model files, thereby reducing the cost of model service construction, improving the efficiency of model application, and enhancing data analysis and processing capabilities.

[0025] In view of this, an embodiment of the present invention provides a method and system for managing the entire life cycle of an aerospace engineering data algorithm model, which is applied to a first device, and the first device communicates with a server. The method includes: storing a plurality of algorithm model files of different categories in a model library; generating an inference flowchart corresponding to a target data analysis process in response to editing operations and configuration operations input by a user in a preset interface; the inference flowchart includes a plurality of nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logical nodes; the editing operation includes adding a node operation, deleting a node operation, and / or connecting a node operation; the configuration operation includes determining the algorithm model file corresponding to each model node from the model library and / or configuring each an execution code corresponding to a logical node, the model node is used to process the input data according to the configured algorithm model file, and the logical node is used to process the input data according to the configured execution code; in response to a publishing operation input by the user in the preset interface, an algorithm model service file corresponding to the target data analysis process is generated according to the inference flowchart and sent to the server, so that the server deploys the algorithm model service according to the algorithm model service file, the algorithm model service is used to respond to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, the data to be analyzed is processed according to the target data analysis process, the data analysis result is obtained, and the data analysis result is sent to the second device.

[0026] The method provided by the present invention stores multiple algorithm model files of different categories in a model library; generates a reasoning flowchart corresponding to the target data analysis process in response to the editing operation and configuration operation input by the user in the preset interface; generates and sends the algorithm model service file corresponding to the target data analysis process to the server according to the reasoning flowchart in response to the publishing operation input by the user in the preset interface, so that the server deploys the algorithm model service according to the algorithm model service file. The method provided by the present invention can quickly complete the deployment of the corresponding algorithm model service based on the analysis configuration operation and the algorithm model file set by the user, and then complete the analysis and processing of the data through the algorithm model service, reduce the cost required for the construction of the model service, improve the efficiency of the model application, and enhance the data analysis and processing capabilities. It can also be understood that: by uniformly managing a variety of heterogeneous model files of different categories, the centralized management and maintenance of the algorithm model files are realized, and the management efficiency is improved. In response to the operation input by the user, the reasoning logic diagram corresponding to the target data analysis process is configured, and the algorithm model service corresponding to the target data analysis process is quickly deployed, which can effectively simplify the deployment process of the service and shorten the cycle from development to deployment of the algorithm model service. In addition, the method provided by the present invention can monitor the running status of the algorithm model service in real time, and ensure the stability and reliability of the model during operation. And the analysis results of the algorithm model service are evaluated through multi-dimensional evaluation indicators, which provides a comprehensive analysis of the performance of the algorithm model service and accurately grasps the performance of the algorithm model service. According to the monitoring and evaluation results, the algorithm model service is iteratively updated through the visualization page, which speeds up the iteration speed of the model and further enhances the performance and adaptability of the model. The method provided by the present invention effectively controls the risks of the model in the production environment through the management of the entire life cycle of the model, improves the comprehensive management level of the model, and ensures the continuity and convenience of the model application.

[0027] See also Figure 1 , Figure 1A schematic diagram of a system architecture is shown in an embodiment of the present invention, including a first device 100, a second device 200 and a server 300. The first device 100 communicates with the server 300, and the second device 200 communicates with the server 300. The first device 100 is used to: store a plurality of different types of algorithm model files in a model library; in response to the editing operation and configuration operation input by the user in the preset interface, generate an inference flowchart corresponding to the target data analysis process; the inference flowchart includes a plurality of nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logic nodes; the editing operation includes a new node operation, a node deletion operation and / or a node connection operation; the configuration operation includes determining the algorithm model file corresponding to each model node from the model library and / or configuring the execution code corresponding to each logic node, the model node is used to process the input data according to the configured algorithm model file, and the logic node is used to process the input data according to the configured execution code; in response to the publishing operation input by the user in the preset interface, generate and send the algorithm model service file corresponding to the target data analysis process to the server according to the inference flowchart. The second device 200 is used to send a data analysis request to the server, wherein the data analysis request carries the data to be analyzed. The server 300 is used to deploy the algorithm model service according to the algorithm model service file, and to respond to the data analysis request sent by the second device 200, process the data to be analyzed according to the target data analysis process, obtain the data analysis result, and send the data analysis result to the second device 200. The second device 200 is also used to receive the data analysis result sent by the server 300.

[0028] In one example, the first device 100, the second device 200 and the server 300 can be any electronic device with data processing capabilities, such as a general-purpose computer, a personal computer, a laptop computer, a switch or a tablet computer, etc. The specific implementation method of the first device 100, the second device 200 and the server 300 is not limited here.

[0029] Figure 2 The electronic device 10 includes a processor 110 , a memory 120 , and a communication interface 130 .

[0030] The processor 110 may include one or more processing cores. The processor 110 uses various interfaces and lines to connect various parts inside the electronic device 10, and executes various functions of the electronic device 10 and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 120, and calling data stored in the memory 120. Optionally, the processor 110 can be implemented in at least one hardware form of a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA).

[0031] The memory 120 may include a random access memory (RAl) or a read-only memory (ROL). Optionally, the memory 120 includes a non-transitory computer-readable storage medium (non-transitory colputer-readable storage lediul). The memory 120 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 120 may include a program storage area. Among them, the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a file storage function, an interface display function, and a model training function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.

[0032] The communication interface 130 is used to communicate with other devices, equipment or communication networks, such as data storage devices, image processing equipment or Ethernet, wireless access network (RAN), wireless local area networks (WLAN), etc. For example, communication between the first device 100 and the server 300, and communication between the second device 200 and the server 300 is realized.

[0033] In physical implementation, the above-mentioned components (such as the processor 110, the memory 120, and the communication interface 130) can be components in the same device (such as a laptop computer). Alternatively, at least two of the components can be set in the same device, that is, as different components in a device, such as a deployment method similar to that of devices or components in a distributed system.

[0034] It is to be understood that the structure illustrated in this embodiment does not constitute a specific limitation on the electronic device 10. In other embodiments of the present invention, the electronic device 10 may include more or fewer components than those illustrated, or combine certain components, or separate certain components, or arrange the components differently. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0035] The following describes a method for managing the entire life cycle of an aerospace engineering data algorithm model provided by an embodiment of the present invention in conjunction with the accompanying drawings.

[0036] Figure 3 The flowchart of a method for managing the entire life cycle of an aerospace engineering data algorithm model provided by an embodiment of the present invention is as follows. Optionally, the method may be Figure 1 The first device 100 shown is executed, that is, Figure 2 The electronic device 10 shown is executed. The method may include the following steps: S1. Store multiple algorithm model files of different categories in the model library.

[0037] Specifically, different categories of algorithm model files are used to implement different data analysis functions and obtain different data analysis results. It should be understood that different categories of algorithm model files are obtained based on different training frameworks, where the training framework is a software library and tools used to build, train and deploy machine learning and deep learning models. The training frameworks include TensorFlow, PyTorch, MLflow, Scikit-learn and Spark-MLlib.

[0038] In this way, the method provided by the embodiment of the present invention can achieve unified management of models with different training frameworks, different categories and different functions, and can solve the compatibility and scalability between models with different training frameworks, different categories and different functions, so that users can flexibly select models with different training frameworks, different categories and different functions for use, thereby effectively improving the deployment efficiency of users' algorithm model services.

[0039] S2. In response to the editing operations and configuration operations input by the user in the preset interface, an inference flowchart corresponding to the target data analysis process is generated. The inference flowchart includes multiple nodes that are configured and connected in sequence according to the target data analysis process. The nodes are model nodes or logic nodes.

[0040] Specifically, the editing operation includes adding a new node, deleting a node, and / or connecting a node; the configuration operation includes determining the algorithm model file corresponding to each model node from the model library and / or configuring the execution code corresponding to each logical node. The model node is used to process the input data according to the configured algorithm model file, and the logical node is used to process the input data according to the configured execution code.

[0041] Among them, the inference flowchart can be understood as a flowchart corresponding to the target data analysis process, and each step in the target data analysis process corresponds to a model node or a logic node. For example, the target data analysis process a is: input the data to be analyzed into the trained algorithm model A, and then input the result output by the algorithm model A into the algorithm model B, and determine the result output by the algorithm model B as the final data analysis result. Then the inference flowchart corresponding to the target data analysis process a includes two model nodes, namely model node A and model node B, wherein the input end of the model node A is connected to the input port for receiving the data to be analyzed and processing the data to be analyzed, and the output end of the model node A is connected to the input end of the model node B for inputting the output result into the model node B, and then the output end of the model node B is connected to the output port for outputting the final data analysis result. Model node A is configured with the algorithm model file corresponding to the algorithm model A, and model node B is configured with the algorithm model file corresponding to the algorithm model B.

[0042] In one possible implementation, the logic node includes a data processing node, a routing node and a combiner node; the data processing node includes at least one input terminal and at least one output terminal, the data processing node is used to process the data input from the input terminal according to the configured execution code, and output the processing result through the output terminal; the combiner node includes at least one input terminal and multiple output terminals, the combiner node is used to output the data input from the input terminal through multiple output terminals respectively according to the configured execution code; the routing node includes at least one input terminal and multiple output terminals, the routing node is used to determine the target output terminal according to the data input from the input terminal through the configured execution code, and output the input data through the target output terminal, and the target output terminal is one of the multiple output terminals.

[0043] For example, see Figure 4 , Figure 4A schematic diagram of a preset interface provided for an embodiment of the present invention. The preset interface includes an operation area and a canvas area. The operation area is used to display model nodes corresponding to different categories of algorithm model files (which can also be understood as displaying models stored in the model library) and logic nodes corresponding to different execution codes. Specifically, the model nodes displayed in the operation area include Scikit-learn models, PyTorch models, TensorFlow models, XGBoost models, MLflow models, PMML models, Triton models, and LightGBM models. The logic nodes displayed in the operation area include data processing nodes, combiner nodes, and routing nodes. The canvas area is used to display the inference flowchart corresponding to the target data analysis process obtained in response to the editing operations and configuration operations input by the user.

[0044] For further information, see Figure 5 , Figure 5 A schematic diagram of an inference flowchart provided for an embodiment of the present invention, the inference flowchart includes 2 data processing nodes, 2 combiner nodes, 1 routing node and 3 model nodes, the 2 data processing nodes are data processing node 1 and data processing node 2, the 2 combiner nodes are combiner node 1 and combiner node 2, and the 3 model nodes are model node 1, model node 2 and model node 3. The input end of the data processing node 1 is used to receive the data to be analyzed, the output end of the data processing node 1 is connected to the input end of the combiner node 1, the two output ends of the combiner node 1 are connected to the input ends of the model node 1 and the routing node, the output end of the routing node is connected to the input ends of the model node 2 and the model node 3, and the output ends of the model node 1, the model node 2 and the model node 3 are used to output the data analysis results of the data to be analyzed. It should be noted that Figure 5 The target data analysis process corresponding to the reasoning flowchart shown is that the data to be analyzed is input into the data processing node 1 through the input port, and then the data processing node 1 processes the data to be analyzed and outputs the processed data to the combiner node 1, and the combiner node 1 outputs the data processed by the data processing node 1 to the model node 1 and the routing node respectively, the model node 1 processes the data processed by the data processing node 1, and outputs the corresponding processing result to the combiner node 2, the routing node judges according to the preset conditions, and outputs the data processed by the data processing node 1 to the model node 2 or the model node 3, and then the model node 2 or the model node 3 processes the data processed by the data processing node 1 and outputs the corresponding processing result to the combiner node 2, the combiner node 2 outputs the processing result output by the output end of the model node 1, the model node 2 and the model node 3 to the data processing node 2, and then the data processing node 2 processes the processing result output by the output end of the model node 1, the model node 2 and the model node 3, and outputs the final data analysis result through the output port.

[0045] It should be understood that the above target data analysis process and the corresponding reasoning flowchart are only exemplary illustrations. The target data analysis process and the corresponding reasoning flowchart provided by the embodiment of the present invention may include more or fewer steps and nodes than the above examples, and the embodiment of the present invention does not impose any special restrictions on this.

[0046] The following is an example to illustrate the specific implementation of the reasoning flowchart provided by the embodiment of the present invention. First, the user adds a model node, a data processing node, a combiner node and a routing node in the canvas area through the new node operation in the preset interface, and then connects the model node, the data processing node, the combiner node and the routing node in sequence according to the target data analysis process through the node connection operation. Among them, the connection of the nodes needs to satisfy that the data processing node is at the top, and the at least two output ends included in the combiner node and the routing node are respectively connected to different model nodes or data processing nodes. Then, in response to the configuration operation input by the user, the corresponding algorithm model file is configured for each model node. The corresponding execution code is configured for each logical node to obtain the reasoning flowchart corresponding to the target data analysis process.

[0047] It can be seen from the above embodiments that the method provided by the present invention can effectively reduce the technical threshold for users to configure complex reasoning processes through a graphical preset interface. Through the graphical drag-and-drop method, the logical branches corresponding to different data analysis processes can be easily configured, and what you see is what you get. Compared with the traditional method of calling multiple interfaces to configure data analysis processes, the method provided by the embodiment of the present invention can effectively simplify the user's operation process and improve the user's experience.

[0048] S3. In response to a publishing operation input by the user in a preset interface, an algorithm model service file corresponding to the target data analysis process is generated according to the inference flowchart and sent to the server, so that the server deploys the algorithm model service according to the algorithm model service file.

[0049] Specifically, the algorithm model service is used to respond to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, processes the data to be analyzed according to the target data analysis process, obtains data analysis results, and sends the data analysis results to the second device.

[0050] Among them, the second device can display the data analysis results in a visual interface, and then carry out subsequent business applications based on the data analysis results.

[0051] It can also be understood as: the algorithm model service is a program running on the server. After the algorithm model service is deployed on the server, the server receives the data analysis request sent by the second device. The server processes the data to be analyzed through the algorithm model service according to the target data analysis process, obtains the data analysis results, and sends the data analysis results to the second device.

[0052] It should be noted that the data to be analyzed may also be obtained from a database storing the data to be analyzed in response to a data analysis request sent by the second device. The embodiment of the present invention does not impose any particular limitation on the specific method of obtaining the data to be analyzed.

[0053] The method provided by the embodiment of the present invention can respond to the user's publishing operation to realize the automatic deployment of the model algorithm service, which can effectively improve the deployment efficiency. The model algorithm service deployment in the related art requires multiple steps such as packaging model files, installing dependent packages, and configuring the server environment, which takes a lot of time and has a complex process. Users are prone to errors during operation, resulting in the failure of model algorithm service deployment. The method provided by the embodiment of the present invention can effectively improve the operation and maintenance efficiency of the model algorithm service.

[0054] In one embodiment, the publishing operation input by the user in the preset interface also carries the service information and resource configuration information input by the user. The service information is used to describe the function of the algorithm model service, so that the user of the second device selects the algorithm model service corresponding to the data analysis process to be used according to the service information, and the resource configuration information is used to describe the resources configured by the algorithm model service, so that the server deploys the algorithm model service based on the resources configured by the resource configuration information.

[0055] In one possible implementation, the algorithm model service is also used to respond to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, adds the data to be analyzed to a task queue arranged in order, processes the multiple data to be analyzed included in the task queue in order according to the target data analysis process, obtains the data analysis result corresponding to each data to be analyzed, and sends the data analysis result corresponding to each data to be analyzed to the second device.

[0056] Specifically, the server adds the data to be analyzed to a task queue arranged in order, and processes the multiple data to be analyzed included in the task queue in order according to the target data analysis process. That is, the server asynchronously calls the algorithm model service in response to the data analysis request sent by the second device, thereby avoiding the inability to process the data to be analyzed that exceeds the resource limit due to the limited resources configured by the algorithm model service, thereby affecting the user experience.

[0057] In a possible implementation, the algorithm model service provided by the embodiment of the present invention supports a model interpretation function, and the server determines the importance score of each of the multiple feature parameters included in the data to be analyzed to the data analysis result, and displays the importance score of each of the multiple feature parameters included in the data to be analyzed to the data analysis result in a visual chart on the second device. Among them, the model interpretation function is applicable to SVM / SVC algorithm, KNN algorithm, logistic regression algorithm, MultinomialNB algorithm and XGBoost algorithm.

[0058] From the above, it can be seen that the method provided by the embodiment of the present invention can effectively enhance the interpretability of the data analysis results of the algorithm model service through the model interpretation algorithm. The user of the second device can view the degree of influence of each feature parameter on the final data analysis results, thereby satisfying the user's usage experience in different usage scenarios and improving the user's usage experience.

[0059] Optionally, the algorithm model service provided by the embodiment of the present invention is configured with a dynamic resource allocation strategy, which can dynamically adjust the number of tasks (instances) to be processed according to the utilization of the central processing unit or memory. This allows the number of instances to be adjusted as the utilization changes. It supports setting the exclusive and shared strategies of the central processing unit.

[0060] In this way, the method provided in the embodiment of the present invention can dynamically adjust the number of execution tasks of the algorithm model service according to the utilization rate of the central processing unit or memory through a dynamic resource configuration strategy, which can effectively save computing resources and improve data processing efficiency.

[0061] As can be seen from the above S1-S3, the method provided by the embodiment of the present invention is to store multiple different categories of algorithm model files in a model library; in response to the editing operation and configuration operation input by the user in the preset interface, generate an inference flowchart corresponding to the target data analysis process; in response to the publishing operation input by the user in the preset interface, generate and send the algorithm model service file corresponding to the target data analysis process to the server according to the inference flowchart, so that the server deploys the algorithm model service according to the algorithm model service file. The method provided by the present invention can quickly complete the deployment of the corresponding algorithm model service based on the analysis configuration operation and algorithm model file set by the user, and then complete the analysis and processing of the data through the algorithm model service, reduce the cost required for model service construction, improve the efficiency of model application, and enhance data analysis and processing capabilities.

[0062] In some embodiments, after the server deploys the algorithm model service according to the algorithm model service file, the method provided by the embodiment of the present invention further includes: An operation status monitoring instruction is sent to the server, and the operation status monitoring instruction is used to instruct: the server obtains the operation status parameters corresponding to the algorithm model service, and the operation status parameters include memory utilization, processor utilization and response time; the server generates a first alarm message when any operation status parameter is greater than or equal to a first threshold and the duration is greater than or equal to a second threshold, and the first alarm message is used to prompt that the operation status of the algorithm model service is abnormal; the server sends the first alarm message to the first device at a preset time interval.

[0063] From the above, it can be seen that the method provided by the embodiment of the present invention can monitor the operating status of the algorithm model service in real time, and when the operating status is abnormal, promptly remind the user of the first device to maintain the deployed algorithm model service, thereby ensuring the stability and reliability of the algorithm model service during operation.

[0064] In some embodiments, after the server deploys the algorithm model service according to the algorithm model service file, the method provided by the embodiment of the present invention further includes: Acquire verification data, where the verification data includes multiple verification samples and a first analysis result obtained by processing each verification sample based on a target data analysis process; An analysis result evaluation instruction is sent to the server, wherein the analysis result evaluation instruction carries verification data, and the analysis result evaluation instruction is used to instruct: the server processes multiple verification samples based on the algorithm model service according to the target data analysis process to obtain a second analysis result corresponding to each verification sample; the server determines the evaluation index corresponding to the algorithm model service according to the first analysis result and the second analysis result corresponding to each verification sample, and the server generates a second alarm message when any evaluation index is greater than or equal to a preset threshold value, and the server sends the second alarm message to the first device, and the second alarm message is used to prompt that the analysis result of the algorithm model service is abnormal; Among them, when the category of the algorithm model service is a binary classification model, the evaluation indicators include: accuracy, precision, recall, specificity, area under the ROC curve, KS test value and F1 measure; when the category of the algorithm model service is a multi-classification model, the evaluation indicators include: accuracy, weighted precision, weighted recall, weighted F value; when the category of the algorithm model service is a regression model, the evaluation indicators include: mean absolute error, mean square error, root mean square error and determination coefficient; when the category of the algorithm model service is a clustering model, the evaluation indicators include: silhouette coefficient-cosine similarity, silhouette coefficient-Euclidean distance, residual sum of squares and covariance.

[0065] Specifically, the ROC curve is the Receiver Operating Characteristic Curve. The KS test value is the Kolmogorov-Smirnov test value.

[0066] It should be understood that the categories of the above algorithm model services and the corresponding evaluation indicators are only exemplary illustrations and are not particularly limited in this regard in the embodiments of the present invention.

[0067] In some embodiments, the first device receives the first analysis result, the second analysis result, the second alarm information, and the evaluation index corresponding to the algorithm model service for each verification sample sent by the server, and displays the first analysis result, the second analysis result, the second alarm information, and the evaluation index corresponding to the algorithm model service for each verification sample on a visual interface.

[0068] In this way, the method provided by the embodiment of the present invention evaluates the data analysis results of the algorithm model service through multi-dimensional evaluation indicators, provides a comprehensive analysis of the performance of the algorithm model service, and accurately grasps the performance of the algorithm model service.

[0069] In some embodiments, see Figure 6 , the method provided by the embodiment of the present invention further includes: S610. In response to an iterative update instruction input by a user, iteratively train the algorithm model files corresponding to one or more model nodes included in the algorithm model service according to the training code to obtain a trained algorithm model file.

[0070] The training code is a pre-configured execution code for iteratively training the algorithm model file corresponding to one or more model nodes included in the algorithm model service according to the training sample. The iterative update instruction carries the preset threshold value of each evaluation indicator in the multiple evaluation indicators corresponding to the algorithm model service, and each evaluation indicator of the algorithm model service configured with the trained algorithm model file is greater than or equal to the preset threshold value.

[0071] S620. Update the algorithm model service file according to the trained algorithm model file to obtain an updated algorithm model service file.

[0072] S630. Send the updated algorithm model service file to the server, so that the server deploys the algorithm model service according to the updated algorithm model service file.

[0073] From the above, it can be seen that the method provided in the embodiment of the present invention enables users to iterate and update instructions to achieve real-time update of the algorithm model service when the algorithm model service does not meet actual usage requirements, thereby ensuring the accuracy and stability of the algorithm model service and meeting the different usage requirements of users in different usage scenarios.

[0074] The following is an example to explain the specific implementation method of iterative update of the algorithm model service provided by an embodiment of the present invention.

[0075] In one example, see Figure 7 The above S610 specifically includes the following steps: S6101. In response to an iterative update instruction input by a user, a continuous training task is created after completing the preceding operations. The continuous training task is used to execute the workflow.

[0076] Specifically, the pre-operation includes obtaining the model training code, completing the model file listing operation, completing the model service deployment operation, and obtaining the model evaluation data set and model evaluation script. The workflow includes multiple configurable nodes, including code nodes, model listing nodes, model evaluation nodes, judgment nodes, and model deployment nodes.

[0077] Among them, the code node is used to perform corresponding operations according to the configured execution code file to realize data preprocessing, feature processing, model training and other links. The judgment node is used to determine whether to continue to execute the next process node according to the custom script. The model launch node is used to perform the model launch operation, and store the algorithm model file generated by the previous node training as a new version of the algorithm model file in the model library. Under the premise of an existing algorithm model service, it is assumed that at least one version of the algorithm model file has been stored in the model library. The model evaluation node is used to perform model evaluation operations on the model nodes that have been deployed and launched. The data that the evaluation task needs to rely on needs to be configured in the database in advance, including evaluation data sets and evaluation scripts. The model deployment node is used to execute the deployment operation of the algorithm model service, and send the updated algorithm model service file to the server as a new service version to the deployed algorithm model service. Under the premise of an already deployed algorithm model service, it is assumed that at least one online version of the algorithm model service has been deployed on the server.

[0078] S6102: Configure parameters corresponding to each configurable node included in the workflow.

[0079] Among them, the parameters of the code node select the instance environment, code file, and model save path used by the above model training code. The parameters of the model launch node select the model that has completed the model launch, and configure the input parameter source as the output parameter of the previous "code node". The parameters of the model evaluation node select the model evaluation dataset and model evaluation script obtained in the above steps, select the evaluation framework as the framework type that the model training depends on, configure the running resources that the model evaluation task needs to rely on, and configure the input parameter source as the output parameter of the previous "model launch" node. The parameters of the judgment node can customize the content of the judgment logic script, such as "judging whether the value of the accuracy indicator in the output result of the previous model evaluation node is greater than the preset threshold", and configure the input parameter source as the output parameter of the previous "model evaluation" node.

[0080] For example, see Figure 8 , Figure 8 A schematic diagram of a node parameter configuration interface provided for an embodiment of the present invention includes a node type display area, a workflow display area, and a parameter setting area. The node type display area is used to display different types of configurable nodes, the workflow area is used to display the workflow corresponding to the continuous training task, and the parameter setting area is used to configure the parameters of each configurable node included in the workflow. The parameters include version information, input parameters, output parameters, and scheduling settings, wherein the scheduling settings include manual triggering and real-time scheduling.

[0081] It should be understood that the above parameters are only exemplary. The configurable nodes provided in the embodiments of the present invention may include more or fewer parameters than the above examples. The embodiments of the present invention do not particularly limit the specific types and quantities of the parameters of the configurable nodes.

[0082] S6103. Execute a continuous training task to iteratively train the algorithm model files corresponding to one or more model nodes included in the algorithm model service according to the training code to obtain a trained algorithm model file.

[0083] In one possible implementation, the continuous training task is configured to perform single or periodic scheduled execution in response to a manual execution instruction or a timed execution instruction input by a user, and to display the task log and evaluation results corresponding to the continuous training task in response to a result display instruction input by a user.

[0084] In order to better illustrate and understand the principle of the method provided by the present invention, the scheme of the present invention is described below in conjunction with an optional specific embodiment. It should be noted that the specific implementation of each step in the specific embodiment should not be understood as a limitation to the scheme of the present invention. On the basis of the principle of the scheme provided by the present invention, other implementations that can be thought of by those skilled in the art should also be considered within the scope of protection of the present invention.

[0085] In this embodiment, if Fig. 9 As shown, a specific embodiment is to perform continuous iterative optimization of the algorithm model service corresponding to the manned space engineering data prediction algorithm model. In this embodiment, the operation is implemented according to the following steps: S710. Store algorithm model files corresponding to different algorithm models in a model library.

[0086] Among them, Table 1 shows the saving methods and file types of algorithm model files supported in the model library.

[0087] Table 1 S720: Obtain incremental data.

[0088] Specifically, write a script file for incremental data reading, configure the task output to be the incremental data file read and the number of incremental data entries, and realize automatic acquisition of incremental data.

[0089] The data used in this embodiment is the telemetry data of manned space engineering. First, based on the defined start and end time, the number of data partitions contained in the time difference is calculated. Combined with the query time point and partition, the SQL query filter condition is constructed to query the incremental data. The result set returned by the query SQL is cached on NFS and stored in the format of a CSV file. The next time the task is run, the last time of the result set will be queried as the start time, and the task running time will be used as the end time.

[0090] S730. Convert the incremental data into a verification data set corresponding to the algorithm model service.

[0091] By configuring the judgment node input, read the output of step S720, judge whether the data meets the model input, and return TRUE or FLASE. If it is TRUE, proceed to the subsequent steps, if it is FLASE, the task is terminated.

[0092] Write a script file for data preprocessing, extract data features, and configure the task to output the preprocessed verification data set.

[0093] S740. Input the verification data set into the algorithm model service to obtain the corresponding data analysis results.

[0094] Write a script file for data prediction, configure the task input as the validation data set output in the previous step, and configure the task output as the data analysis result. Write a script file for evaluation index calculation, configure the task input as the validation data set output in the previous step and the data analysis result, and configure the task output as the model evaluation index. The evaluation indicators of different types of algorithm models are described in the above embodiments, which will not be repeated here.

[0095] The validation data set and data analysis results are saved in two ways: NFS storage and database storage. NFS storage writes data to a fixed file directory. After the task is run, the result file can be downloaded from the interface. Database storage reads the database through a custom function and queries the data by passing in the database name, table name, and column name.

[0096] S750. Based on the evaluation index of the algorithm model service, determine whether to retrain the algorithm model files included in the algorithm model service, and iterate the algorithm model files stored in the model library.

[0097] By configuring the judgment node, read the evaluation indicators output in the previous step, and judge whether the evaluation indicators meet the preset conditions. If the model indicators do not meet the preset conditions, regenerate the verification data set, train the algorithm model corresponding to the algorithm model file, and return TRUE or FLASE. If it is TRUE, proceed to the subsequent steps, if it is FLASE, the task is terminated.

[0098] Write a script file for model training. The task output is configured as the new version of the algorithm model file produced by the model training task. The new version of the algorithm model file is stored in the model library. The name of the new version of the algorithm model file is set based on the storage time.

[0099] S760. When the algorithm model file stored in the model library is iterated according to the trained algorithm model file, the deployed algorithm model service is updated based on the iterated algorithm model file.

[0100] It should be understood that the server can deploy multiple different versions of algorithm model services at the same time. When receiving a data analysis request sent by the second device, the algorithm model service determines the algorithm model file to be called according to the pre-set publishing mode and outputs the corresponding data analysis results. For a detailed description of the publishing mode, please refer to the following embodiments.

[0101] The method of the present invention solves the problems in the prior art that various heterogeneous models are difficult to manage in a unified manner, the model deployment cycle is long, and the value of the model cannot be fully released due to the lack of a unified standardized process. The algorithm model is provided with full life cycle management of unified management, unified operation and maintenance, unified application, unified monitoring, and unified evaluation, thereby achieving easy-to-use, efficient, safe and reliable model operation management.

[0102] In some embodiments, the method provided by the embodiment of the present invention further includes: A data distribution difference monitoring instruction is sent to the server, wherein the data distribution difference monitoring instruction carries training data for each algorithm model file included in the algorithm model service, and the data distribution difference monitoring instruction is used to indicate: when the server receives a data analysis request sent by the second device, it determines the data distribution difference value between the data to be analyzed and the training data carried in the data analysis request, and when the data distribution difference value is greater than or equal to a third threshold, the server generates a third alarm message, and the server sends the third alarm message to the second device, and the third alarm message is used to prompt that there is a data distribution difference between the data to be analyzed and the training data.

[0103] The embodiment of the present invention monitors the data distribution difference value between the data to be analyzed and the training data, and can adaptively adjust the training data and retrain to obtain an algorithm model service that meets the actual application scenario, thereby preventing users from obtaining data analysis results with poor accuracy, thereby improving the stability of the algorithm model service.

[0104] In some embodiments, when a server is deployed with multiple algorithm model services corresponding to the target data analysis process, the method provided by the embodiment of the present invention further includes: Sending a release mode setting instruction to the server, the release mode setting instruction is used to instruct: the server to set the release mode of multiple algorithm model services corresponding to the target data analysis process; wherein the release mode includes independent release, grayscale release or comparative release; When the publishing mode is independent publishing, the publishing mode setting instruction carries an identifier of a target algorithm model service, the target algorithm model service is one of multiple algorithm model services, the server responds to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, the server processes the data to be analyzed through the target algorithm model service, obtains a data analysis result, and sends the data analysis result to the second device; When the publishing mode is grayscale publishing, the server responds to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, the server processes the data to be analyzed through multiple algorithm model services to obtain the data analysis results corresponding to each algorithm model service, and sends the data analysis results corresponding to each algorithm model service to the second device; When the publishing mode is comparative publishing, the server responds to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, the server processes the data to be analyzed through multiple algorithm model services to obtain the data analysis results corresponding to each algorithm model service, the server determines the evaluation index corresponding to each algorithm model service, and the server sends the data analysis result of the algorithm model service corresponding to the maximum value of the evaluation index corresponding to each algorithm model service to the second device.

[0105] Specifically, the evaluation metrics are positively correlated with the performance of the algorithm model service.

[0106] In some embodiments, when the evaluation index is negatively correlated with the performance of the algorithm model service, when the publishing mode is comparative publishing, the server sends the data analysis results of the algorithm model service corresponding to the minimum value of the evaluation index corresponding to each algorithm model service to the second device.

[0107] It should be understood that the publishing mode of the algorithm model service is sent to the server through the algorithm model service file corresponding to the algorithm model service before the server deploys the algorithm model service. The embodiment of the present invention does not impose any special restrictions on the specific setting method of the publishing mode of the algorithm model service.

[0108] From the above, it can be seen that the method provided by the embodiment of the present invention can flexibly set different publishing modes of the algorithm model service. When the server deploys multiple algorithm model services corresponding to the target data analysis process (which can also be understood as different versions of the same algorithm model service), it can be based on different publishing modes. Use different versions of the same algorithm model service to analyze the data to be analyzed and obtain data analysis results, thereby meeting the user's usage needs in different usage scenarios.

[0109] The above mainly introduces the scheme of the embodiment of the present invention from the perspective of the method. It is understandable that in order to realize the above functions, the first device 100 includes at least one of the hardware structure and software modules corresponding to the execution of each function. It should be easily appreciated by those skilled in the art that, in combination with the units and algorithm steps of each example described in the embodiments disclosed herein, the embodiments of the present invention can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the embodiments of the present invention.

[0110] The embodiment of the present invention can divide the first device 100 into functional units according to the above method example. For example, the first device 100 can be divided into functional units corresponding to various functions, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. It should be noted that the division of units in the embodiment of the present invention is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0111] For example, Fig.10A schematic diagram of the structure of a full life cycle management system for aerospace engineering data algorithm model provided by an embodiment of the present invention is shown. The management system 400 is applied to a first device, and the first device communicates with a server, and includes: a model management module 410, which is used to store multiple algorithm model files of different categories in a model library; a service configuration module 420, which is used to generate an inference flowchart corresponding to a target data analysis process in response to editing operations and configuration operations input by a user in a preset interface; the inference flowchart includes multiple nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logical nodes; editing operations include adding nodes, deleting nodes, and / or connecting nodes; configuration operations include determining the algorithm model file corresponding to each model node from the model library and / or configuring the execution code corresponding to each logical node. The model node is used to process the input data according to the configured algorithm model file, and the logic node is used to process the input data according to the configured execution code; the service publishing module 430 is used to respond to the publishing operation input by the user in the preset interface, generate and send the algorithm model service file corresponding to the target data analysis process to the server according to the inference flowchart, so that the server deploys the algorithm model service according to the algorithm model service file, and the algorithm model service is used to respond to the data analysis request sent by the second device, the data analysis request carries the data to be analyzed, processes the data to be analyzed according to the target data analysis process, obtains the data analysis result, and sends the data analysis result to the second device.

[0112] Optionally, the management system 400 also includes an operation status monitoring module 440, which is used to: send an operation status monitoring instruction to the server, the operation status monitoring instruction is used to instruct: the server to obtain the operation status parameters corresponding to the algorithm model service, the operation status parameters include memory utilization, processor utilization and response time; the server generates a first alarm message when any operation status parameter is greater than or equal to a first threshold and the duration is greater than or equal to a second threshold, the first alarm message is used to prompt that the operation status of the algorithm model service is abnormal; the server sends the first alarm message to the first device at a preset time interval.

[0113] Optionally, the management system 400 also includes an analysis result evaluation module 450, which is used to: obtain verification data, the verification data including multiple verification samples and a first analysis result obtained by each verification sample based on the target data analysis process; send an analysis result evaluation instruction to the server, the analysis result evaluation instruction carries the verification data, and the analysis result evaluation instruction is used to instruct: the server processes the multiple verification samples according to the target data analysis process based on the algorithm model service to obtain a second analysis result corresponding to each verification sample; the server determines the evaluation index corresponding to the algorithm model service based on the first analysis result and the second analysis result corresponding to each verification sample, and generates a second alarm message when any evaluation index of the server is greater than or equal to a preset threshold, and the server The server sends a second alarm message to the first device, and the second alarm message is used to prompt that the analysis result of the algorithm model service is abnormal; wherein, when the category of the algorithm model service is a binary classification model, the evaluation indicators include: accuracy, precision, recall, specificity, area under the ROC curve, KS test value and F1 measure; when the category of the algorithm model service is a multi-classification model, the evaluation indicators include: accuracy, weighted precision, weighted recall, weighted F value; when the category of the algorithm model service is a regression model, the evaluation indicators include: mean absolute error, mean square error, root mean square error and determination coefficient; when the category of the algorithm model service is a clustering model, the evaluation indicators include: silhouette coefficient-cosine similarity, silhouette coefficient-Euclidean distance, residual sum of squares and covariance.

[0114] Optionally, the management system 400 also includes a service iteration update module 460, which is used to: respond to an iteration update instruction input by a user, iteratively train the algorithm model files corresponding to one or more model nodes included in the algorithm model service according to the training code to obtain a trained algorithm model file; update the algorithm model service file according to the trained algorithm model file to obtain an updated algorithm model service file; send the updated algorithm model service file to the server, so that the server deploys the algorithm model service according to the updated algorithm model service file.

[0115] It should be understood that the specific description of the above optional methods can refer to the above method embodiments, which will not be repeated here. In addition, the explanation of any management system 400 provided above and the description of the beneficial effects can refer to the above corresponding method embodiments, which will not be repeated here.

[0116] The embodiment of the present invention further provides a computer-readable storage medium, in which at least one computer instruction is stored, and the at least one computer instruction is loaded and executed by a processor to implement the methods of the above embodiments. For the explanation of the relevant contents and the description of the beneficial effects in any of the above-mentioned computer-readable storage media, reference can be made to the above-mentioned corresponding embodiments, which will not be repeated here.

[0117] The embodiment of the present invention further provides a chip. The chip integrates a control circuit and one or more ports for implementing the functions of the management system 400. Optionally, the functions supported by the chip can be referred to above and will not be described in detail here.

[0118] Those skilled in the art will appreciate that all or part of the steps of the above embodiments can be implemented by instructing the relevant hardware through a program, and the program can be stored in a computer-readable storage medium. The above-mentioned storage medium can be a read-only memory, a random access memory, etc. The above-mentioned processing unit or processor can be a central processing unit, a general-purpose processor, a specific circuit structure (application specific integrated circuit, ASIC), a microprocessor (digital signal processor, DSP), a field programmable gate array (field prograllable gatearray, FPGA) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof.

[0119] The embodiment of the present invention also provides a computer program product including instructions, which, when executed on a computer, enables the computer to perform any of the methods in the above embodiments. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more available media integrated therein. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., an SSD), etc.

[0120] It should be noted that the above-mentioned devices for storing computer instructions or computer programs provided in the embodiments of the present invention, such as but not limited to the above-mentioned memories, computer-readable storage media, and communication chips, etc., are all non-transitory. Those skilled in the art should be aware that in one or more of the above examples, the functions described in the embodiments of the present invention can be implemented by hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable storage medium or transmitted as one or more instructions or codes on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein the communication medium includes any medium that facilitates the transmission of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0121] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A method for managing the entire life cycle of an aerospace engineering data algorithm model, characterized in that: Applied to a first device, the first device communicates with a server, the method comprising: Store multiple different categories of algorithm model files in the model library; In response to the editing operation and configuration operation input by the user in the preset interface, an inference flowchart corresponding to the target data analysis process is generated; the inference flowchart includes a plurality of nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logic nodes; the editing operation includes a new node operation, a node deletion operation and / or a node connection operation; the configuration operation includes determining the algorithm model file corresponding to each model node from the model library and / or configuring the execution code corresponding to each logic node, the model node is used to process the input data according to the configured algorithm model file, and the logic node is used to process the input data according to the configured execution code; In response to a publishing operation input by the user in a preset interface, an algorithm model service file corresponding to the target data analysis process is generated according to the inference flowchart and sent to the server, so that the server deploys an algorithm model service according to the algorithm model service file, and the algorithm model service is used to respond to a data analysis request sent by a second device, the data analysis request carries data to be analyzed, processes the data to be analyzed according to the target data analysis process, obtains a data analysis result, and sends the data analysis result to the second device.

2. The method according to claim 1, characterized in that The logic node includes a data processing node, a routing node and a combiner node; the data processing node includes at least one input terminal and at least one output terminal, and the data processing node is used to process the data input from the input terminal according to the configured execution code, and output the processing result through the output terminal; the combiner node includes at least one input terminal and multiple output terminals, and the combiner node is used to output the data input from the input terminal through multiple output terminals respectively according to the configured execution code; The routing node includes at least one input terminal and multiple output terminals. The routing node is used to determine a target output terminal through a configured execution code according to data input from the input terminal, and output the input data through the target output terminal, where the target output terminal is one of the multiple output terminals.

3. The method according to claim 2, characterized in that After the server deploys the algorithm model service according to the algorithm model service file, the method further includes: Send an operation status monitoring instruction to the server, wherein the operation status monitoring instruction is used to instruct: the server to obtain the operation status parameters corresponding to the algorithm model service, wherein the operation status parameters include memory utilization, processor utilization and response time; the server generates a first alarm message when any operation status parameter is greater than or equal to a first threshold and the duration is greater than or equal to a second threshold, and the first alarm message is used to prompt that the operation status of the algorithm model service is abnormal; the server sends the first alarm message to the first device at a preset time interval.

4. The method according to claim 3, characterized in that After the server deploys the algorithm model service according to the algorithm model service file, the method further includes: Acquire verification data, the verification data including a plurality of verification samples and a first analysis result obtained by processing each verification sample based on the target data analysis process; An analysis result evaluation instruction is sent to the server, wherein the analysis result evaluation instruction carries the verification data, and the analysis result evaluation instruction is used to instruct: the server processes the multiple verification samples based on the algorithm model service according to the target data analysis process to obtain a second analysis result corresponding to each verification sample; the server determines the evaluation index corresponding to the algorithm model service according to the first analysis result and the second analysis result corresponding to each verification sample, and the server generates a second alarm message when any evaluation index is greater than or equal to a preset threshold value, and the server sends the second alarm message to the first device, and the second alarm message is used to prompt that the analysis result of the algorithm model service is abnormal; Among them, when the category of the algorithm model service is a binary classification model, the evaluation indicators include: accuracy, precision, recall, specificity, area under the ROC curve, KS test value and F1 measure; when the category of the algorithm model service is a multi-classification model, the evaluation indicators include: accuracy, weighted precision, weighted recall, weighted F value; when the category of the algorithm model service is a regression model, the evaluation indicators include: mean absolute error, mean square error, root mean square error and determination coefficient; when the category of the algorithm model service is a clustering model, the evaluation indicators include: silhouette coefficient-cosine similarity, silhouette coefficient-Euclidean distance, residual sum of squares and covariance.

5. The method according to claim 4, characterized in that The method further comprises: In response to an iterative update instruction input by a user, iteratively train the algorithm model files corresponding to one or more model nodes included in the algorithm model service according to the training code to obtain a trained algorithm model file; The algorithm model service file is updated according to the trained algorithm model file to obtain an updated algorithm model service file; The updated algorithm model service file is sent to the server so that the server deploys the algorithm model service according to the updated algorithm model service file.

6. The method according to claim 5, characterized in that The method further comprises: A data distribution difference monitoring instruction is sent to the server, wherein the data distribution difference monitoring instruction carries training data for each algorithm model file included in the algorithm model service, and the data distribution difference monitoring instruction is used to instruct: when the server receives the data analysis request sent by the second device, it determines the data distribution difference value between the data to be analyzed and the training data carried in the data analysis request, and when the data distribution difference value is greater than or equal to a third threshold, the server generates a third alarm message, and the server sends the third alarm message to the second device, and the third alarm message is used to prompt that there is a data distribution difference between the data to be analyzed and the training data.

7. The method according to claim 6, characterized in that The algorithm model service is also used to respond to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, adds the data to be analyzed to a task queue arranged in order, processes the multiple data to be analyzed included in the task queue in order according to the target data analysis process, obtains the data analysis result corresponding to each data to be analyzed, and sends the data analysis result corresponding to each data to be analyzed to the second device.

8. The method according to claim 7, characterized in that In the case where the server is deployed with a plurality of algorithm model services corresponding to the target data analysis process, the method further includes: Sending a publishing mode setting instruction to the server, wherein the publishing mode setting instruction is used to instruct: the server to set a publishing mode of multiple algorithm model services corresponding to the target data analysis process; wherein the publishing mode includes independent publishing, grayscale publishing or comparative publishing; In the case where the publishing mode is independent publishing, the publishing mode setting instruction carries an identifier of a target algorithm model service, the target algorithm model service is one of a plurality of algorithm model services, the server responds to a data analysis request sent by a second device, the data analysis request carries data to be analyzed, the server processes the data to be analyzed through the target algorithm model service, obtains a data analysis result, and sends the data analysis result to the second device; In the case where the release mode is grayscale release, the server responds to a data analysis request sent by the second device, the data analysis request carries data to be analyzed, the server processes the data to be analyzed through multiple algorithm model services to obtain a data analysis result corresponding to each algorithm model service, and sends the data analysis result corresponding to each algorithm model service to the second device; When the publishing mode is comparative publishing, the server responds to a data analysis request sent by the second device, wherein the data analysis request carries data to be analyzed, and the server processes the data to be analyzed through multiple algorithm model services to obtain data analysis results corresponding to each algorithm model service. The server determines the evaluation index corresponding to each algorithm model service, and the server sends the data analysis result of the algorithm model service corresponding to the maximum value of the evaluation index corresponding to each algorithm model service to the second device.

9. A full life cycle management system for aerospace engineering data algorithm model, characterized in that: Applied to a first device, the first device communicates with a server, and the system includes: A model management module is used to store multiple algorithm model files of different categories in a model library; A service configuration module, for generating a reasoning flowchart corresponding to a target data analysis process in response to an editing operation and a configuration operation input by a user in a preset interface; the reasoning flowchart includes a plurality of nodes that are configured and connected in sequence according to the target data analysis process, and the nodes are model nodes or logic nodes; the editing operation includes a new node operation, a node deletion operation, and / or a node connection operation; the configuration operation includes determining an algorithm model file corresponding to each model node from the model library and / or configuring an execution code corresponding to each logic node, the model node is used to process the input data according to the configured algorithm model file, and the logic node is used to process the input data according to the configured execution code; A service publishing module is used to respond to a publishing operation input by a user in a preset interface, generate an algorithm model service file corresponding to the target data analysis process according to the inference flowchart, and send it to the server, so that the server deploys an algorithm model service according to the algorithm model service file, and the algorithm model service is used to respond to a data analysis request sent by a second device, the data analysis request carries data to be analyzed, processes the data to be analyzed according to the target data analysis process, obtains a data analysis result, and sends the data analysis result to the second device.

10. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; Wherein, the processor is configured to execute the instructions to implement the aerospace engineering data algorithm model full life cycle management method as described in any one of claims 1-8.

Images