Multi-task process integrated collaborative design method

Abstract

The embodiment of the invention discloses a multi-task process integrated collaborative design method, which belongs to the technical field of test tasks, and comprises the following steps: defining task steps, acquiring decomposition granularity, decomposing a test task according to the decomposition granularity and a task relationship, and acquiring sub-tasks and meta-tasks, the method comprises the following steps: describing and decomposing a three-dimensional structure of a task according to the Hall three-dimensional structure, constructing a Hall management matrix, constructing a task model according to the Hall management matrix, concluding and classifying the test task according to the task model, generating a complete test task flow, and simplifying and verifying the test task flow. According to the embodiment of the invention, the subtask sequence which can be directly executed can be obtained, meanwhile, the distinguishing degree between the meta-tasks is high, the task granularity is prevented from being too fine, and unnecessary computing resources and time cannot be wasted.

Description

Technical Field

[0001] The present application belongs to the field of experimental task technology, and specifically, relates to a multi-task process integrated collaborative design method. Background Art

[0002] The formulation of an operational test plan is the first step in conducting an operational test. Traditional unmanned system tests are mainly based on performance test evaluation. The operational test scenarios and operational test elements are relatively simple, the test boundary constraints are relatively clear, and the test process is basically based on linear testing. Manual test demand analysis, test planning, and program design can basically meet the needs of sequential implementation, organizational scheduling, and resource allocation of traditional tests.

[0003] At present, the preparation of test plans before tests has always relied mainly on manual labor, which to some extent limits the efficiency and quality of combat tests and has the following problems: manual preparation of test plans takes a long time. The preparation of test plans usually needs to consider multiple variables and parameters, and perform multiple optimizations and adjustments. The traditional manual preparation method requires experts to spend a lot of time and energy on analysis and design, which makes the process of test plan preparation very slow; human bias leads to the plan being not scientific and reasonable enough. Manually formulated test plans are easily affected by individual experience, subjective consciousness and cognitive bias, resulting in the plan being not scientific and reasonable enough. There is a lack of systematic analysis and optimization, and the test plan may not fully explore and discover the essence of the problem; it is difficult to meet the needs of complex tests. With the advancement of science and technology and the expansion of application fields, the complexity of tests is also increasing. Test plans need to consider more factors and constraints. Manually formulated test plans are often unable to meet the needs of complex tests; there is a lack of standardization and normalization in manually formulated test plans. There may be differences in the plan formulation methods between different test personnel and institutions, which affects the comparability and repeatability of the test results. Summary of the invention

[0004] In order to solve the above problems and technical defects, the embodiment of the present application adopts the following technical solution, a multi-task process integrated collaborative design method, comprising the following steps:

[0005] Step 1: Define the task steps, obtain the decomposition granularity, decompose the test task according to the decomposition granularity and task relationship, and obtain subtasks and meta-tasks;

[0006] Step 2: Describe and decompose the three-dimensional structure of the task according to the Hall three-dimensional structure, construct the Hall management matrix, and construct the task model according to the Hall management matrix;

[0007] Step 3: Summarize and classify the test tasks according to the task model, generate a complete test task process, and then simplify and verify the test task process.

[0008] Preferably, the process of constructing the Hall management matrix is ​​as follows:

[0009] S01. Determine the contents of the time dimension and the logic dimension in the three-dimensional structure, combine the task phase represented by the time dimension and the experimental link represented by the logic dimension, and establish the Hall management primary matrix;

[0010] The three-dimensional results include: time dimension, logic dimension, and activity dimension;

[0011] The task phases of the time dimension include: test preparation phase, test implementation phase, and test end phase;

[0012] The experimental stages of the logical dimension include: observation, judgment, decision-making, and execution;

[0013] S02, decompose the test tasks and fill them into the Hall management primary matrix;

[0014] If the element in the Hall management primary matrix does not involve a subtask, it is determined that the subtask corresponding to the element does not exist. If the element in the Hall management primary matrix involves multiple subtasks, it is determined that the subtask corresponding to the element can be further decomposed.

[0015] S03. Clarify the test time stage and test logic link of each subtask, and obtain the subtask set;

[0016] S04, judging whether the subtask can be further decomposed according to the subtask set, if it cannot be decomposed, the subtask can be defined as a meta-task and proceed to the next step; otherwise, go to step S02;

[0017] S05. Standardize the description of meta-tasks so that meta-tasks only contain descriptions of three dimensions: time, logic, and activity, and establish the Hall management matrix.

[0018] Furthermore, the standardized description of the meta-task is a description of the meta-task attributes and relationships, completing the description of the task list and the task network. The task list is a collection of meta-tasks, describing the meta-task attributes themselves; the task network is a collection of meta-task relationships, describing the various relationships between meta-tasks. The standardized description of meta-tasks includes: standardized description of meta-task attributes and standardized description of meta-task relationships.

[0019] Furthermore, the meta-task relationship normalization description includes meta-task temporal relationship, meta-task logical relationship and meta-task functional relationship.

[0020] Furthermore, the meta-task timing relationship is defined as seven relationships in the time dimension of the test tasks according to the order of each meta-task in the time dimension during the implementation of the test tasks, using interval algebra as the definition of the workflow time relationship. The meta-task timing relationship includes: serial connection relationship, serial interval relationship, serial overlapping relationship, parallel inclusion relationship, parallel start relationship, parallel end relationship, and parallel equality relationship.

[0021] Furthermore, the meta-task logical relationship refers to the logical relationship between each meta-task except the temporal relationship, and the meta-task logical relationship includes: AND relationship, AND-OR relationship, OR-AND relationship, OR-OR relationship, and loop relationship.

[0022] Furthermore, the meta-task functional relationship refers to the functional relationship between each meta-task, and the meta-task functional relationship includes: dependency relationship, coupling relationship, and independent relationship.

[0023] Preferably, the Hall management matrix constructs the task model by adopting a particle swarm clustering optimization algorithm based on multi-task collaboration to obtain the optimal solution for each task, combining the optimal solution with expert knowledge to obtain the clustering result of each data set, and using the clustering external indicator to adjust the Rand coefficient ARI to evaluate the quality of the clustering results, and using the clustering result with a qualified quality evaluation result as the task model.

[0024] Furthermore, the quality evaluation process is as follows:

[0025] The actual partition of the N samples in the data set is represented as P, the partition obtained by clustering the algorithm is Q, and the two different samples of the data set are divided into X i and X i , the two samples are divided into four different situations:

[0026] X i and X j Belonging to the same cluster in Q, they also belong to the same category in P;

[0027] X i and X j Belong to the same cluster in Q, but belong to different categories in P;

[0028] X i and X j Belong to different clusters in Q, belong to the same category in P;

[0029] X i and X j They belong to different clusters in Q and do not belong to the same category in P;

[0030] The number of samples of the four situation pairs is counted and expressed as a, b, c, and d. The ARI calculation formula is as follows:

[0031]

[0032] When ARI is 1, it indicates that the two partitions are consistent. A comparison threshold is preset. If ARI is greater than or equal to the comparison threshold, it indicates that the clustering effect evaluation is qualified. If ARI is less than the comparison threshold, it indicates that the clustering effect evaluation is unqualified.

[0033] Furthermore, the Pi calculation rule is used to simplify the test task process, and the formula is as follows:

[0034] P::=A <a1,a2,…,a n >|∑ i∈I π i ·P i |(P1|P2)|P1+P2|[a=b]P|new ap|! P|0

[0035] Among them, A <a1,a2,...,a n > indicates that the test task process A has parameters inside, and the names of these parameters are a1, a2, ..., a n ;∑ i∈I π i ·P i represents a sum-type expression, where P i π i Guarded by π i After the action is executed, process P i to start the activity; P1|P2 represents a parallel expression, where process P1 and process P2 can be executed concurrently, share the name between channels, and communicate through channels; P1+P2 represents a selection expression, where process P1 and process P2 can only choose one to execute, and the other process will be abandoned; [a=b]P represents a matching expression, where a and b represent names, and process P will only execute when a is equal to b. new ap represents a restriction expression, where name a is restricted in process P; ! P represents a copy expression, which is equivalent to multiple processes P executing concurrently. The number of processes P can be dynamically created as needed, and new process expressions are defined by recursion; 0 represents an empty process expression, where 0 can be omitted.

[0036] Compared with the prior art, the beneficial effects of the embodiments of the present application are:

[0037] This application is approved. The test task is decomposed layer by layer, and detailed decomposition and description are completed. Each meta-task obtained after decomposition should be able to fully describe the original experimental task. The test task is divided into three dimensions of "time-logic-activity" for decomposition, which solves the problem of large granularity of meta-tasks obtained when directly referring to the decomposition of test activities, and is competent for the needs of complex experiments. As the decomposition process proceeds, the granularity of the task gradually decreases, and the content contained is more specific and richer, the essence of the problem is discovered, and a sub-task sequence that can be directly executed is obtained. At the same time, there is a high degree of distinction between meta-tasks, which also avoids the task granularity being too fine, and does not waste unnecessary computing resources and time. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In the attached picture:

[0039] Figure 1 A schematic diagram of the method steps of an embodiment of the present application;

[0040] Figure 2 This is a framework diagram of the test task decomposition based on the Hall three-dimensional structure of an embodiment of the present application. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Generally, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0042] Example

[0043] like Figure 1 As shown, a multi-task process integrated collaborative design method includes the following steps:

[0044] Define the task steps, obtain the decomposition granularity, decompose the test task according to the decomposition granularity and task relationship, and obtain subtasks and meta-tasks;

[0045] The test mission is determined by the commander based on the battlefield situation, the superior's intention, the test style and other test factors, and is issued in the form of orders or instructions. It is usually macro and vague.

[0046] Subtasks can be further decomposed into other subtasks or metatasks;

[0047] Meta-task refers to a sub-task that cannot be further decomposed or the granularity reaches the required level. It can be directly executed by the test unit and is the most basic task that constitutes the test task.

[0048] Decomposition granularity is a measure of the degree of refinement of task decomposition, and is generally expressed by the coarsest granularity in the meta-task set. If the granularity is too coarse, the task decomposition process will be inadequate and the units will be difficult to understand; if the granularity is too fine, unnecessary computing resources and time will be wasted, and even delays will occur.

[0049] Task relationship,During the execution of the experimental task, there are complex interactive relationships,among the meta-tasks. The input of some meta-tasks often comes from other meta-tasks, and the execution process depends on,other tasks for support. This collaborative relationship between tasks is called,task relationship.

[0050] Describe and decompose the tasks in three dimensions according to Hall's three-dimensional structure, and construct the Hall management matrix;

[0051] The process of building the Hall management matrix is ​​as follows:

[0052] S01. Determine the contents of the time dimension and the logic dimension in the three-dimensional structure, combine the task phase represented by the time dimension and the experimental link represented by the logic dimension, and establish the Hall management primary matrix;

[0053] The three-dimensional results include: time dimension, logic dimension, and activity dimension;

[0054] The task phases of the time dimension include: test preparation phase, test implementation phase, and test end phase;

[0055] The experimental stages of the logical dimension include: observation, judgment, decision-making, and execution;

[0056] S02, decompose the test tasks and fill them into the Hall management primary matrix;

[0057] If the element in the Hall management primary matrix does not involve a subtask, it is determined that the subtask corresponding to the element does not exist. If the element in the Hall management primary matrix involves multiple subtasks, it is determined that the subtask corresponding to the element can be further decomposed.

[0058] Stage dimension\action dimension observe judge decision making implement Test preparation stage <![CDATA[a 11 ]]> <![CDATA[a 12 ]]> <![CDATA[a 13 ]]> <![CDATA[ a14 ]]> Trial and Experimental Phase <![CDATA[a 21 ]]> <![CDATA[a 22 ]]> <![CDATA[a 23 ]]> <![CDATA[a 24 ]]> End of trial <![CDATA[a 31 ]]> <![CDATA[a 31 ]]> <![CDATA[ a33 ]]> <![CDATA[ a34 ]]>

[0059] For example, at the end of the experiment, no execution is required, that is, the "End of the experiment-Execution" element a 34 There are no subtasks;

[0060] The trial implementation phase requires multiple judgments, and the "trial implementation phase - judgment" element a 22 The corresponding subtasks can be decomposed;

[0061] S03. Clarify the test time stage and test logic link of each subtask, and obtain the subtask set;

[0062] S04, judging whether the subtask can be further decomposed according to the subtask set, if it cannot be decomposed, the subtask can be defined as a meta-task and proceed to the next step; otherwise, go to step S02;

[0063] S05. Standardize the description of meta-tasks so that meta-tasks only contain descriptions of three dimensions: time, logic, and activity, and establish the Hall management matrix.

[0064] The standardized description of meta-tasks is a description of the meta-task attributes and relationships, completing the description of the task list and task network. The task list is a collection of meta-tasks, describing the meta-task attributes themselves. The task network is a collection of meta-task relationships, describing the various relationships between meta-tasks. The standardized description of meta-tasks includes: standardized description of meta-task attributes and standardized description of meta-task relationships.

[0065] The task list is expressed as: T = <t1,t2,…,t i ,…,t n >

[0066] Among them, t i is the i-th meta-task, n is the number of meta-tasks;

[0067] The task network representation formula is: R = <r1,r2,…,r i ,…,r m >

[0068] Among them, r i is the i-th meta-task relationship, and m is the number of meta-task relationships.

[0069] The calculation formula for the normalized description of meta-task attributes is as follows:

[0070] T d = <T type , T time , T space , T subject , T index >

[0071] Among them, T type is the task type, T time is the task time, T space is the task space, T subject is the task subject, T index For task indicators;

[0072] Task Type T type It refers to the classification of test tasks at the logical level. The calculation formula is as follows:

[0073] T type = <P type , P name , P id >

[0074] Among them, P type is the type name, which is divided into four categories: intelligence collection, intelligence processing, command and control, response execution, name is the task name, which is an abstract simplification of the experimental task description. id The task number is the unique identifier for distinguishing tasks;

[0075] Task time T time It is the key to describing the timing relationship of tasks and the embodiment of the execution time of the test task. The calculation formula is as follows:

[0076] T time = <T sta , T end , T con >

[0077] Among them, T sta is the task start time, T end is the task end time, T con is the duration of the task;

[0078] Task Space T space It refers to the actual geographical location of the mission target. The calculation formula is as follows:

[0079] T space = type , S xy >

[0080] Among them, S type is the space type where the target is located, including various test spaces such as land, water, underwater, air, and space. xy For the convenience of demonstration, the space is simplified into a two-dimensional plane and the target coordinates are simplified into two-dimensional coordinates. However, the target coordinates are very complex in the actual process. The reference system adopted will be different depending on the experimental background.

[0081] Task Body T subject Refers to the main equipment and forces that perform the test mission. The calculation formula is as follows:

[0082] T subject = name , U number >

[0083] Among them, U name is the name of the task execution entity, U number The number of tasks required.

[0084] Task index T index ​​It refers to the demand situation of the task on the ability index of the execution subject, which is expressed in the form of demand vector. The calculation formula is as follows:

[0085] T inde x= <i1,i2,…,i n >

[0086] Among them, i n is the nth capability indicator.

[0087] The standardized description of meta-task relations includes meta-task temporal relations, meta-task logical relations and meta-task functional relations;

[0088] The meta-task timing relationship is defined as seven relationships in the time dimension of the test task during the implementation of the test task, based on the order of each meta-task in the time dimension, using interval algebra as the definition of the time relationship of the workflow, including: serial connection relationship, serial interval relationship, serial overlapping relationship, parallel inclusion relationship, parallel start relationship, parallel end relationship, and parallel equality relationship.

[0089] The relational expression of the serial connection relationship is r meet (t i , t j ), represents the meta-task t i and t j It is carried out in chronological order without any interval;

[0090] The relational expression for the serial interval relation is r before (t i , t j ), represents the meta-task t i and t j It is performed in chronological order after an interval of d seconds:

[0091] The relational expression of the serial overlap relation is r overlaps (t i , t j ), represents the meta-task t i and t j It is carried out in chronological order. i d seconds before the end of t j Start the task;

[0092] The relational expression of the parallel inclusion relation is r during (t i , t j ), represents the meta-task t i and t j are carried out in the same period, and t i and t j The start and end times are different;

[0093] The relational expression of the parallel start relation is r starts (t i , t j ), represents the meta-task t i and t j are carried out in the same period, and t i and t j The start time is the same;

[0094] The relational expression for the parallel end relation is r finishes (t i , t j ), represents the meta-task t i and t j are carried out in the same period, and t i and t j The end time is the same;

[0095] The relational expression for the parallel equality relation is r equal (t i , t j ), represents the meta-task t i and t j are carried out in the same period, and t i and t j The start and end times are the same.

[0096] Meta-task logical relationships refer to the logical relationships between each meta-task in addition to the temporal relationship. For example, some meta-tasks have corresponding predecessor tasks, so the logical level also needs to be considered when constructing meta-task relationships.

[0097] Meta-task logical relations include: AND-and relation, AND-OR relation, OR-AND relation, OR-OR relation, and loop relation;

[0098] And the relationship, for T1, T2∈T, there exists a relationship r and_and (T1, T2), which means that all tasks in task set T2 can be started only when all tasks in task set T1 are completed;

[0099] AND or relation, for T1, T2∈T, |T2|>2, there exists relation r and_or (T1, T2), which means that only when all tasks in task set T1 are completed, some tasks in task set T2 can be started;

[0100] Or and relation, for T1, T2∈T, |T1|≥2, there exists relation r or_and (T1, T2), which means that when some tasks in task set T1 are completed, all tasks in task set T2 can start to be executed;

[0101] Or or relation, for T1, T2∈T, |T1|≥2V|T2|≥2, there exists a relation r or_or (T1, T2), which means that only when some tasks in task set T1 are completed, some tasks in task set T2 can be started;

[0102] Cyclic relationship, for t i ∈T, there exists a relation r loop (t i ), indicating that when the meta-task t i After completion, the subsequent tasks remain for themselves and fall into a single-task loop.

[0103] Meta-task functional relationship refers to the functional relationship between each meta-task. For example, the execution of some meta-tasks requires information or resource support from other tasks. Therefore, when constructing meta-task relationships, the functional relationship level needs to be considered. Meta-task functional relationships include: dependency relationship, coupling relationship, and independent relationship.

[0104] Dependency, for t i , t j ∈T, there exists a relation r dependent (t i , t j ), represents the meta-task t j Need to rely on t i Provide information, substance or effect to execute;

[0105] Coupling relationship, for t i , t j ∈T, there exists a relation r couple (t i , t j ), represents the meta-task t i and t j Relying on information, substances or effects provided by each other for execution;

[0106] Independence relationship, for t i , t j ∈T, if the meta-task t i and t j If each party can execute without any support from the other party, it is called t i and t j The two are independent and are represented by r independent (t i , t j ).

[0107] Constructing a task model based on the Hall Management Matrix;

[0108] The Hall management matrix constructs the task model by using a particle swarm clustering optimization algorithm based on multi-task collaboration to obtain the optimal solution for each task, and then combines the optimal solution with expert knowledge to obtain the clustering result of each data set. The clustering external indicator is used to adjust the Rand coefficient ARI to evaluate the quality of the clustering results, and the clustering results with qualified quality evaluation results are used as the task model.

[0109] ARI is an improved index of the proposed clustering evaluation coefficient Rand Index, which makes the index close to zero when the clustering results are randomly generated.

[0110] The quality evaluation process is as follows:

[0111] The actual partition of the N samples in the data set is represented as P, the partition obtained by clustering the algorithm is Q, and the two different samples of the data set are divided into X i and X j , the two samples are divided into four different situations:

[0112] X i and X j They belong to the same cluster in Q and they also belong to the same category in P;

[0113] X i and X j belong to the same cluster in Q, but they belong to different categories in P;

[0114] X i and X j They belong to different clusters in Q, but they belong to the same category in P;

[0115] X i and X j They belong to different clusters in Q, and they do not belong to the same category in P;

[0116] The number of samples of the four situation pairs is counted and expressed as a, b, c, and d. The ARI calculation formula is as follows:

[0117]

[0118] When ARI is 1, it indicates that the two partitions are consistent. A comparison threshold is preset. If ARI is greater than or equal to the comparison threshold, it indicates that the clustering effect evaluation is qualified. If ARI is less than the comparison threshold, it indicates that the clustering effect evaluation is unqualified.

[0119] The test tasks are summarized and classified according to the task model to generate a complete test task process, which is then simplified and verified.

[0120] The Pi calculation rule is used to simplify the test task process. The formula is as follows:

[0121] P::=A <a1,a2,…,a n >|∑ i∈I π i ·P i |(P1|P2)|P1+P2|[a=b]P|new ap|! P|0

[0122] Among them, A <a1,a2,...,a n > indicates that the test task process A has parameters inside, and the names of these parameters are a1, a2, ..., a n ;

[0123] ∑ i∈I π i ·P i represents a sum-type expression, where P i π i Guarded by π i After the action is executed, process P i To start the activity;

[0124] P1|P2 represents a parallel expression, where processes P1 and P2 can be executed concurrently, share the channel name, and communicate through the channel;

[0125] P1+P2 represents a selection expression, where only one of the processes P1 and P2 can be executed, and the other process will be abandoned;

[0126] [a=b]P represents a matching expression, where a and b represent names. Process P will be executed only when a is equal to b.

[0127] new ap represents a restricted expression, in which the name a is confined in process P. It is worth noting that the confined name a in process P and the free name a outside process P can exist at the same time, but their roles and meanings are completely different. The confined name a is only meaningful in process P, while the free name a is meaningful in the entire process communication.

[0128] ! P represents a replication expression, which is equivalent to multiple processes P executing concurrently. The number of processes P can be created dynamically as needed, and new process expressions are defined recursively;

[0129] 0 represents an empty process expression, where 0 can be omitted.

[0130] The above-mentioned embodiments only express the preferred implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for ordinary technicians in this field, several modifications, improvements and substitutions can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.

Claims

1. A multi-task process integrated collaborative design method, characterized in that: The following steps are involved: Step 1: Define the task steps, obtain the decomposition granularity, decompose the test task according to the decomposition granularity and task relationship, and obtain subtasks and meta-tasks; Step 2: Describe and decompose the three-dimensional structure of the task according to the Hall three-dimensional structure, construct the Hall management matrix, and construct the task model according to the Hall management matrix; Step 3: Summarize and classify the test tasks according to the task model, generate a complete test task process, and then simplify and verify the test task process.

2. A multi-task process integrated collaborative design method according to claim 1, characterized in that: The process of constructing the Hall management matrix is ​​as follows: S01. Determine the contents of the time dimension and the logic dimension in the three-dimensional structure, combine the task phase represented by the time dimension and the experimental link represented by the logic dimension, and establish the Hall management primary matrix; The three-dimensional results include: time dimension, logic dimension, and activity dimension; The task phases of the time dimension include: test preparation phase, test implementation phase, and test end phase; The experimental stages of the logical dimension include: observation, judgment, decision-making, and execution; S02, decompose the test tasks and fill them into the Hall management primary matrix; If the element in the Hall management primary matrix does not involve a subtask, it is determined that the subtask corresponding to the element does not exist. If the element in the Hall management primary matrix involves multiple subtasks, it is determined that the subtask corresponding to the element can be further decomposed. S03. Clarify the test time stage and test logic link of each subtask, and obtain the subtask set; S04, judging whether the subtask can be further decomposed according to the subtask set, if it cannot be decomposed, the subtask can be defined as a meta-task and proceed to the next step; otherwise, go to step S02; S05. Standardize the description of meta-tasks so that meta-tasks only contain descriptions of three dimensions: time, logic, and activity, and establish the Hall management matrix.

3. A multi-task process integrated collaborative design method according to claim 2, characterized in that: The standardized description of the meta-task is a description of the meta-task attributes and relationships, completing the description of the task list and the task network. The task list is a collection of meta-tasks, describing the meta-task attributes themselves. The task network is a collection of meta-task relationships, describing the various relationships between meta-tasks. The standardized description of meta-tasks includes: standardized description of meta-task attributes and standardized description of meta-task relationships.

4. A multi-task process integrated collaborative design method according to claim 3, characterized in that: The meta-task relationship normalization description includes meta-task temporal relationship, meta-task logical relationship and meta-task functional relationship.

5. A multi-task process integrated collaborative design method according to claim 4, characterized in that: The meta-task timing relationship is defined as seven relationships in the time dimension of the test tasks according to the order of each meta-task in the time dimension during the implementation of the test tasks, using interval algebra as the definition of the workflow time relationship. The meta-task timing relationship includes: serial connection relationship, serial interval relationship, serial overlapping relationship, parallel inclusion relationship, parallel start relationship, parallel end relationship, and parallel equality relationship.

6. A multi-task process integrated collaborative design method according to claim 4, characterized in that: The meta-task logical relationship refers to the logical relationship between each meta-task except the temporal relationship. The meta-task logical relationship includes: AND relationship, AND-OR relationship, OR-AND relationship, OR-OR relationship, and loop relationship.

7. A multi-task process integrated collaborative design method according to claim 4, characterized in that: The meta-task functional relationship refers to the functional relationship between each meta-task, and the meta-task functional relationship includes: dependency relationship, coupling relationship, and independent relationship.

8. The multi-task process integrated collaborative design method according to claim 1, characterized in that: The Hall management matrix constructs the task model by adopting a particle swarm clustering optimization algorithm based on multi-task collaboration to obtain the optimal solution for each task, combining the optimal solution with expert knowledge to obtain the clustering result of each data set, and using the clustering external indicator to adjust the Rand coefficient ARI to evaluate the quality of the clustering result, and taking the clustering result with a qualified quality evaluation result as the task model.

9. A multi-task process integrated collaborative design method according to claim 8, characterized in that: The quality evaluation process is as follows: The actual partition of the N samples in the data set is represented as P, the partition obtained by clustering the algorithm is Q, and the two different samples of the data set are divided into X i and X j , the two samples are divided into four different situations: X i and X j Belonging to the same cluster in Q, they also belong to the same category in P; X i and X j Belong to the same cluster in Q, but belong to different categories in P; X i and X j Belong to different clusters in Q, belong to the same category in P; X i and X j They belong to different clusters in Q and do not belong to the same category in P; The number of samples of the four situation pairs is counted and expressed as a, b, c, and d. The ARI calculation formula is as follows: When ARI is 1, it indicates that the two partitions are consistent. A comparison threshold is preset. If ARI is greater than or equal to the comparison threshold, it indicates that the clustering effect evaluation is qualified. If ARI is less than the comparison threshold, it indicates that the clustering effect evaluation is unqualified.

10. The multi-task process integrated collaborative design method according to claim 1, characterized in that: The Pi calculation rule is used to simplify the test task process, and the formula is as follows: P::=A <a1,a2,...,a n >|∑ i∈I π i ·P i |(P1|P2)|P1+P2|[a=b]P|new ap|!P|0 Among them, A <a1,a2,...,a n > indicates that the test task process A has parameters inside, and the names of these parameters are a1, a2, ..., a n ;∑ i∈I π i ·P i represents a sum-type expression, where P i π i Guarded by π i After the action of i to start the activity; P1|P2 represents a parallel expression, where process P1 and process P2 can be executed concurrently, share the name between channels, and communicate through channels; P1+P2 represents a selection expression, where process P1 and process P2 can only choose one to execute, and the other process will be abandoned; [a=b]P represents a matching expression, where a and b represent names, and process P will only execute when a is equal to b. new ap represents a restriction expression, where name a is restricted in process P; ! P represents a copy expression, which is equivalent to multiple processes P executing concurrently. The number of processes P can be dynamically created as needed, and new process expressions are defined by recursion; 0 represents an empty process expression, where 0 can be omitted.

Images