A multi-project test full life cycle integration management system and method
By integrating and managing the entire lifecycle of multi-project testing, the system solves the problems of uneven resource distribution and isolated test cases in multi-project collaboration scenarios of traditional test management systems. It enables dynamic resource allocation, intelligent test case matching, and cross-project risk warning, thereby improving testing efficiency and quality.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHANGHAI KENGNU INFORMATION TECHNOLOGY CO LTD
- Filing Date
- 2026-01-26
- Publication Date
- 2026-06-19
AI Technical Summary
Traditional test management systems suffer from uneven resource management, isolated test case libraries, and a lack of cross-project risk discovery and early warning capabilities when facing multi-project collaboration scenarios, resulting in low testing efficiency and difficulty in ensuring quality.
It employs a multi-project dynamic access module, an AI intelligent resource scheduling module, a cross-project use case reuse module, a blockchain evidence storage module, and a cross-project risk warning module to achieve dynamic resource allocation, intelligent use case matching, distributed evidence storage, and cross-project risk warning.
It enables intelligent coordination and dynamic optimization of testing resources across multiple projects, improving testing efficiency and coverage, ensuring the credibility and transparency of the testing process, and providing high-quality project delivery assurance.
Smart Images
Figure CN122240458A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of software test management technology, and more specifically, to a multi-project test full lifecycle integrated management system and method. Background Technology
[0002] With the large-scale and intensive development of the software R&D industry, it has become the norm for enterprises to simultaneously promote multiple parallel testing projects. The testing tasks cover the entire lifecycle from requirement analysis, test case design, execution scheduling, defect tracking to report generation, and must meet core requirements such as cross-team collaboration, resource sharing, and quality traceability.
[0003] Traditional test management systems, when faced with complex multi-project collaborative scenarios, rely heavily on static strategies or manual experience for resource allocation. This makes it difficult to dynamically respond to changes in project priorities, real-time loads, and task complexity, leading to uneven resource utilization and inter-project conflicts, thus impacting overall testing efficiency. Secondly, regarding test asset reuse, the lack of intelligent test case retrieval and matching mechanisms means test case libraries often become information silos, failing to effectively mine and utilize the value of historical test cases, resulting in extensive repetitive design and hindering the improvement of test coverage and the accumulation of test assets. Finally, in terms of process quality management, reliable evidence storage mechanisms for key test node data are generally lacking, and defect analysis is mostly limited to single-project implementations, lacking the ability to uncover and warn of cross-project related risks, making it difficult to achieve reliable traceability and proactive risk prevention throughout the entire process. Summary of the Invention
[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a multi-project testing full life cycle integrated management system and method to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, the present invention provides the following technical solution: a multi-project testing full lifecycle integrated management system, comprising: The multi-project dynamic access module is used to receive the requirement information, priority parameters and resource requests of at least two parallel test projects. The full lifecycle integration management module communicates with the multi-project dynamic access module and is used to integrate the entire process management of test requirement analysis, test case design, test execution, defect closure and report generation. The AI intelligent resource scheduling module communicates with the multi-project dynamic access module and the full lifecycle integration management module respectively, and is used to dynamically allocate test resources based on multi-project priority, resource load status and test task complexity. The cross-project test case reuse module communicates with the full lifecycle integration management module and is used to achieve intelligent retrieval, matching and iterative optimization of test cases for multiple projects based on machine learning models; The blockchain evidence storage module communicates with the full lifecycle integrated management module and is used to generate digital fingerprints for test key node data and perform distributed evidence storage. The cross-project risk early warning module communicates with the full lifecycle integrated management module and the defect closed-loop unit to mine cross-project defect correlations and trigger dynamic early warnings.
[0006] Preferably, the priority evaluation model of the multi-project dynamic access module adopts the analytic hierarchy process, and the priority parameters include project urgency, business value weight, resource consumption coefficient, and delivery deadline threshold.
[0007] Preferably, the AI intelligent resource scheduling module adopts the Gradient Boosting Tree (GBRT) algorithm, and the input features include project priority score, current resource load rate, test task complexity coefficient and historical scheduling success rate, and the output resource allocation scheme and conflict resolution strategy.
[0008] Preferably, the cross-project test case reuse module achieves intelligent retrieval, matching, and iterative optimization of test cases through the following methods: Extract the attribute information of the test cases, including function type, input parameters, expected results, preconditions and applicable scenario tags; Based on the extracted attributes, a spectral clustering algorithm is used to cluster historical use cases from multiple projects to form use case clusters; A reinforcement learning model is adopted, with the improvement of test case reuse success rate and test coverage as reward functions, and the test case matching algorithm is iteratively optimized.
[0009] Preferably, the blockchain evidence storage module is configured as follows: The first digital fingerprint is generated based on the SM3 algorithm, and the second digital fingerprint is generated based on the SHA-256 algorithm. The first digital fingerprint and the second digital fingerprint are hashed again to generate a fused digital fingerprint, and the fused digital fingerprint is synchronized to the consortium blockchain node. The key node data includes test case versions, test execution records, defect reports, and acceptance reports.
[0010] It also provides a method for integrated management of the entire lifecycle of multi-project testing, including the following steps: S1: Input parallel project information through the multi-project dynamic access module and complete priority evaluation based on the analytic hierarchy process; S2: The AI intelligent resource scheduling module generates a resource allocation scheme based on priority scores, resource load, and task complexity using a gradient boosting tree algorithm; S3: The full lifecycle integration management module initiates the testing process, and the cross-project test case reuse module retrieves and optimizes available test cases through spectral clustering and reinforcement learning models. S4: During the test execution process, the blockchain evidence storage module generates dual digital fingerprints for the test key node data and stores them in a distributed manner. The key node data includes test case execution records and defect data. S5: The cross-project risk warning module mines defect association rules and triggers a warning when the density of associated defects exceeds a dynamic threshold. S6: After the defect is closed, a multi-project test report is generated, and the test case reuse library and resource scheduling model parameters are updated synchronously.
[0011] Preferably, the matching operation in S3 is implemented in the following way: the similarity of test cases is calculated by using a weighted fusion algorithm of cosine similarity and Jaccard similarity, and the weight coefficients of the algorithm are iteratively optimized through a reinforcement learning model.
[0012] Preferably, the key node data in S4 also includes test environment configuration parameters, resource allocation logs, and cross-project collaboration records, and the consortium blockchain nodes include project party, test party, and supervision party nodes.
[0013] Preferably, the cross-project risk warning module in S5 uses the Apriori algorithm to mine defect association rules, and the dynamic threshold is dynamically adjusted according to the historical defect transmission rate and project priority.
[0014] Preferably, in S6: The use case reuse library update includes clustering newly added use cases, adjusting the weights of historical use cases based on reuse performance, and identifying use cases with low reuse rates. The resource scheduling model parameter update cycle is synchronized with the project iteration cycle.
[0015] The technical effects and advantages of this invention are as follows: By introducing a priority evaluation model based on the analytic hierarchy process (AHP) and an AI-powered intelligent resource scheduling module using the gradient boosting tree algorithm, the system can comprehensively consider multiple parameters such as project urgency, business value, and resource consumption. Based on real-time load and historical scheduling success rate, it can dynamically and accurately allocate resources and resolve conflicts, overcoming the shortcomings of traditional manual scheduling, which is characterized by strong subjectivity and delayed response. This enables a shift from passive response to proactive prediction, thereby maximizing resource utilization efficiency in a multi-project parallel environment, shortening the overall testing cycle, and achieving intelligent coordination and dynamic optimization of multi-project testing resources, significantly improving testing efficiency and resource utilization. By intelligently classifying historical test cases using spectral clustering algorithms and continuously optimizing the matching algorithm with reuse success rate and test coverage as reward functions using reinforcement learning models, accurate retrieval and intelligent recommendation of test cases are achieved. This not only uncovers implicit test case associations and avoids redundant design, but also continuously improves matching accuracy through weighted fusion of cosine similarity and Jaccard similarity, thereby forming a continuous accumulation and virtuous cycle of enterprise test assets, ensuring test quality from the source, and building a cross-project intelligent test case reuse mechanism with self-learning capabilities, effectively improving test coverage and test case design quality. By integrating blockchain notarization and data mining technologies, and generating a fused digital fingerprint using SM3 and SHA-256 dual hashing and synchronizing it to the consortium blockchain, the immutability and traceability of data at key testing nodes are ensured. At the same time, defect association rules are mined based on the Apriori algorithm, and the warning threshold is dynamically adjusted according to historical transmission rate and project priority, enabling early detection and precise prevention of potential risks. An integrated quality assurance system that combines full-process credible traceability and cross-project risk forward-looking early warning has been established, greatly enhancing the credibility and transparency of the testing process and providing a solid guarantee for the high-quality delivery of complex projects. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the system structure of the present invention.
[0017] Figure 2 This is a schematic diagram of the method flow structure of the present invention. Detailed Implementation
[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0019] This invention relates to the field of software test management technology, and in particular to a multi-project test full lifecycle integrated management system and method.
[0020] As attached Figure 1 As shown, a multi-project test lifecycle integrated management system is deployed on a server, which includes a central processing unit (CPU), RAM, and storage such as a hard disk. The core of the system consists of the following modules: The multi-project dynamic access module is used to receive the requirement information, priority parameters and resource requests of at least two parallel test projects. The priority evaluation model for the multi-project dynamic access module adopts the analytic hierarchy process (AHP). The judgment matrix is constructed by pairwise comparison scaling and the eigenvector is calculated to determine the weight of each parameter. The priority parameters include project urgency, business value weight, resource consumption coefficient, and delivery deadline threshold. The full lifecycle integration management module communicates and connects with the multi-project dynamic access module to integrate the entire process management of test requirement parsing, test case design, test execution, defect closure and report generation. This module provides a RESTful API interface for each module to call and uses message queues such as Kafka to achieve asynchronous communication between modules. The AI intelligent resource scheduling module communicates with the multi-project dynamic access module and the full lifecycle integration management module respectively, and is used to dynamically allocate test resources based on the priority of multiple projects, resource load status and test task complexity. The AI intelligent resource scheduling module adopts the gradient boosting tree (GBRT) algorithm. This model uses historical project data, including project features and resource usage records, for supervised learning training to minimize the error between the predicted allocation scheme and the actual optimal scheme. Input features include project priority score, current resource load rate, test task complexity coefficient, and historical scheduling success rate. Output resource allocation scheme and conflict resolution strategy. The cross-project test case reuse module communicates with the full lifecycle integration management module to achieve intelligent retrieval, matching, and iterative optimization of test cases across multiple projects based on machine learning models. The cross-project test case reuse module achieves intelligent retrieval, matching, and iterative optimization of test cases through the following methods: Extract the attribute information of the test cases, including function type, input parameters, expected results, preconditions and applicable scenario labels; the attribute information is vectorized and used as model input; Based on the extracted attributes, a spectral clustering algorithm is used to cluster historical use cases of multiple projects to form use case clusters. Specifically, a use case similarity matrix is first constructed, then feature decomposition is performed to reduce dimensionality, and finally the K-means algorithm is used to cluster the feature vectors. A reinforcement learning model is adopted, with the improvement of test case reuse success rate and test coverage as the reward function, and the test case matching algorithm is iteratively optimized. The reinforcement learning model adopts DQN, which is a deep Q-network algorithm. Its state space is the test case features and its action space is the matching policy parameters. The blockchain evidence storage module communicates with the full lifecycle integration management module to generate digital fingerprints for key test node data and perform distributed evidence storage. The blockchain evidence storage module is configured as follows: The first digital fingerprint is generated based on the SM3 algorithm, and the second digital fingerprint is generated based on the SHA-256 algorithm. After concatenating the first digital fingerprint with the second digital fingerprint, the SM3 algorithm is used again for hashing to generate a fused digital fingerprint, which is then synchronized to the consortium blockchain node. The consortium blockchain is built using the Hyperledger Fabric framework, and the evidence data is stored in JSON format. Key node data includes test case version, test execution record, defect report and acceptance report; The cross-project risk warning module communicates with the full lifecycle integration management module and the defect closed-loop unit to mine cross-project defect correlations and trigger dynamic warnings. This module monitors the data flow of defect management platforms such as Jira in real time, and sends warning information via email or instant messaging interface when it finds related defect patterns.
[0021] The system is also connected to databases such as MySQL to store test case libraries, project resource libraries, defect libraries, and model parameters.
[0022] A method for integrated management of the entire lifecycle of multi-project testing; As attached Figure 2 As shown, the method includes the following steps: S1: Parallel project information is entered through the multi-project dynamic access module, and priority evaluation is completed based on the analytic hierarchy process. The specific evaluation process includes: the project manager and the test manager compare each parameter pairwise according to the scaling method, generate a judgment matrix, and calculate the eigenvector corresponding to the largest eigenvalue as the weight vector. Finally, the project priority score is calculated by weighting. S2: The AI intelligent resource scheduling module generates a resource allocation scheme based on priority scores, resource load, and task complexity using the gradient boosting tree algorithm; the GBRT model is implemented using the Scikit-learn library, and hyperparameters are determined through grid search. S3: The full lifecycle integration management module initiates the testing process, and the cross-project test case reuse module retrieves and optimizes available test cases through spectral clustering and reinforcement learning models. In S3, the matching operation is implemented as follows: the similarity of test cases is calculated by using a weighted fusion algorithm of cosine similarity and Jaccard similarity, and the weight coefficients of this algorithm are iteratively optimized through a reinforcement learning model; the initial weights are set to 0.5:0.5, and the reinforcement learning model adjusts the weights according to the test results of each reuse, such as passing or failing. S4: During test execution, the blockchain evidence storage module generates dual digital fingerprints for key test node data and stores them in a distributed manner. Key node data includes test case execution records and defect data. Key node data in S4 also includes test environment configuration parameters, resource allocation logs, and cross-project collaboration records. Consortium blockchain nodes include project party, test party, and supervision party nodes. Each node can synchronize and verify the stored information only after it has been authenticated with a digital certificate. S5: The cross-project risk warning module mines defect association rules and triggers a warning when the density of associated defects exceeds a dynamic threshold. In S5, the cross-project risk warning module uses the Apriori algorithm to mine defect association rules, with a minimum support of 0.1 and a minimum confidence of 0.7. The dynamic threshold is dynamically adjusted based on the historical defect transmission rate and project priority. The adjustment formula is: threshold = baseline threshold × (1 + project priority coefficient) × (1 + historical transmission rate). S6: After the defect is closed, a multi-project test report is generated, and the test case reuse library and resource scheduling model parameters are updated synchronously. In S6: The use case reuse library update includes clustering new use cases, adjusting the weight of historical use cases based on reuse performance, and identifying use cases with low reuse rates. Use cases with low reuse rates refer to use cases that have not been used for five consecutive iteration cycles, and the system marks them as needing optimization. The resource scheduling model parameter update cycle is synchronized with the project iteration cycle; after each project iteration, the data generated in this iteration is used for incremental model training.
[0023] The method also includes displaying test progress, resource heatmaps and risk warning panels in real time through a web front-end interface, and all operation logs are recorded in the system log file for auditing.
[0024] Finally, it should be noted that the accompanying drawings of the embodiments disclosed in this invention only involve the structures involved in the embodiments disclosed in this invention. Other structures can refer to the general design. In the absence of conflict, the same embodiment and different embodiments of this invention can be combined with each other. In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A multi-project testing full lifecycle integrated management system, characterized in that: The multi-project dynamic access module is used to receive the requirement information, priority parameters and resource requests of at least two parallel test projects. The full lifecycle integration management module communicates with the multi-project dynamic access module and is used to integrate the entire process management of test requirement analysis, test case design, test execution, defect closure and report generation. The AI intelligent resource scheduling module is connected to the multi-project dynamic access module and the full lifecycle integration management module, respectively, and is used to dynamically allocate test resources based on multi-project priority, resource load status and test task complexity. The cross-project test case reuse module communicates with the full lifecycle integration management module and is used to achieve intelligent retrieval, matching and iterative optimization of test cases for multiple projects based on machine learning models. The blockchain evidence storage module is connected to the full lifecycle integrated management module and is used to generate digital fingerprints for test key node data and perform distributed evidence storage. The cross-project risk early warning module is communicatively connected to the full lifecycle integrated management module and the defect closed-loop unit, and is used to explore the cross-project defect correlation and trigger dynamic early warning.
2. The multi-project test full lifecycle integrated management system according to claim 1, characterized in that: The priority evaluation model of the multi-project dynamic access module adopts the analytic hierarchy process, and the priority parameters include project urgency, business value weight, resource consumption coefficient, and delivery deadline threshold.
3. The multi-project test full lifecycle integrated management system according to claim 1, characterized in that: The AI intelligent resource scheduling module adopts the gradient boosting tree algorithm. The input features include project priority score, current resource load rate, test task complexity coefficient and historical scheduling success rate. The output is a resource allocation scheme and conflict resolution strategy.
4. The multi-project test full lifecycle integrated management system according to claim 1, characterized in that: The cross-project test case reuse module achieves intelligent retrieval, matching, and iterative optimization of test cases through the following methods: Extract the attribute information of the test cases, including function type, input parameters, expected results, preconditions and applicable scenario tags; Based on the extracted attributes, a spectral clustering algorithm is used to cluster historical use cases from multiple projects to form use case clusters; A reinforcement learning model is adopted, with the improvement of test case reuse success rate and test coverage as reward functions, and the test case matching algorithm is iteratively optimized.
5. The multi-project test full lifecycle integrated management system according to claim 1, characterized in that: The blockchain evidence storage module is configured as follows: The first digital fingerprint is generated based on the SM3 algorithm, and the second digital fingerprint is generated based on the SHA-256 algorithm. The first digital fingerprint and the second digital fingerprint are hashed again to generate a fused digital fingerprint, and the fused digital fingerprint is synchronized to the consortium blockchain node. The key node data includes test case versions, test execution records, defect reports, and acceptance reports.
6. A method for integrated management of the entire lifecycle of multi-project testing, characterized in that: Includes the following steps: S1: Input parallel project information through the multi-project dynamic access module and complete priority evaluation based on the analytic hierarchy process; S2: The AI intelligent resource scheduling module generates a resource allocation scheme based on priority scores, resource load, and task complexity using a gradient boosting tree algorithm; S3: The full lifecycle integration management module initiates the testing process, and the cross-project test case reuse module retrieves and optimizes available test cases through spectral clustering and reinforcement learning models. S4: During the test execution process, the blockchain evidence storage module generates dual digital fingerprints for the test key node data and stores them in a distributed manner. The key node data includes test case execution records and defect data. S5: The cross-project risk warning module mines defect association rules and triggers a warning when the density of associated defects exceeds a dynamic threshold. S6: After the defect is closed, a multi-project test report is generated, and the test case reuse library and resource scheduling model parameters are updated synchronously.
7. The multi-project testing full lifecycle integrated management method according to claim 6, characterized in that: The matching operation in S3 is implemented as follows: the similarity of test cases is calculated by using a weighted fusion algorithm of cosine similarity and Jaccard similarity, and the weight coefficients of the algorithm are iteratively optimized through a reinforcement learning model.
8. The multi-project testing full lifecycle integrated management method according to claim 6, characterized in that: The key node data in S4 also includes test environment configuration parameters, resource allocation logs, and cross-project collaboration records. The consortium blockchain nodes include project party, test party, and supervision party nodes.
9. The multi-project testing full lifecycle integrated management method according to claim 6, characterized in that: The cross-project risk warning module in S5 uses the Apriori algorithm to mine defect association rules, and the dynamic threshold is dynamically adjusted according to the historical defect transmission rate and project priority.
10. The multi-project testing full lifecycle integrated management method according to claim 6, characterized in that: In S6: The use case reuse library update includes clustering newly added use cases, adjusting the weights of historical use cases based on reuse performance, and identifying use cases with low reuse rates. The resource scheduling model parameter update cycle is synchronized with the project iteration cycle.