A model-based embedded software quality quantification control method

Abstract

The application discloses a model-based embedded software quality quantitative control method, comprising the following steps: (1) calculating defect detection target values of each process segment according to an organization performance baseline; (2) completing curve fitting of a strain variable Y and an influence factor X; (3) calculating a recommended review rate or a use case density by using the fitted curve according to a process detection target number, a review scale or a test scale of a project; and (4) performing Monte Carlo simulation on a probability of achieving the detection target number of the test process. The application can give recommended influence factor values, such as a recommended review rate and a recommended use case density value, by using the fitted curve of the control target and the influence factor based on the given control target, so as to control effective execution of the review and test processes of a new project, thereby guaranteeing software development quality.

Description

Technical Field

[0001] The present invention relates to quality quantification management, in particular to a model-based embedded software quality quantification control method. Background Art

[0002] According to the definition of CMMI (Capability Maturity Model Integration), the maturity level of software development is divided into five levels: Initial, Managed, Defined, Quantitatively Managed, and Optimized. Maturity level is an important evaluation that reflects the continuous evolution and improvement of organizational processes. At level LV4 (Quantitatively Managed), software development organizations are required to have the ability to analyze detailed measurement data of software processes and product quality and to quantitatively understand and control software processes and products. Currently, most software organizations are at level LV3 or have some initial capabilities of level LV4, but they have not been able to fully explain how to measure software processes and product quality through data and implement corresponding controls. Summary of the Invention

[0003] Purpose of the invention: The purpose of the present invention is to provide a model-based embedded software quality quantitative control method, which can be used to guide the project to control the review rate during the review and ensure a reasonable use case density or test period during system testing to ensure the quality of software research and development.

[0004] Technical solution: The model-based embedded software quality quantitative control method described in the present invention includes the following steps:

[0005] (1) Calculate the defect detection target value for each process segment based on the organizational performance baseline;

[0006] (2) Complete the curve fitting of strain Y and influencing factor X;

[0007] (3) Based on the project's process checkout target number, review scale, or test scale, a fitting curve is used to calculate the recommended review rate or use case density;

[0008] (4) Perform Monte Carlo simulation on the probability of achieving the target number of detections during the test process.

[0009] The step (1) comprises the following steps:

[0010] (1.1) Complete statistical analysis of the performance baseline of defect implantation detection in each process segment;

[0011] (1.2) Calculation of estimated defect implantation values ​​for each process segment;

[0012] (1.3) Calculation of defect detection target value for each process segment.

[0013] The step (2) comprises the following steps:

[0014] (2.1) Complete the curve fitting of the number of requirements detected and the rate of requirements review;

[0015] (2.2) Complete the curve fitting of the number of code review checkouts and the code review rate;

[0016] (2.3) Complete the curve fitting of the number of system test defects detected, test case density, and test process progress pressure.

[0017] The step (3) comprises the following steps:

[0018] (3.1) Based on the target number of defects detected in the requirements process and the scale of the requirements review, the recommended requirements review rate is calculated using the fitted requirements detection number and requirements review rate curve to be used for quality control of the requirements review process;

[0019] (3.2) Based on the target number of defects detected during the design implementation process and the scale of code review, a fitted curve of the number of defects detected during the design implementation process and the code review rate is used to calculate the recommended code review rate for quality control of the code review process.

[0020] (3.3) Based on the target number of defects detected during the test process and the scale of the lines of code tested, the recommended test case density is calculated using the fitted test process detection number and test case density, and test process progress pressure rate curve for quality control of the system test process.

[0021] The step (4) comprises the following steps:

[0022] (4.1) During the planning phase, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detections using estimated lines of code;

[0023] (4.2) Before transitioning from development to testing, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detections after the revision using actual code lines.

[0024] A computer storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned model-based embedded software quality quantitative control method.

[0025] A computer device includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, the above-mentioned model-based embedded software quality quantitative control method is implemented.

[0026] Beneficial effects: Compared with the existing technology, the present invention has the following advantages: by fitting the control target and the impact factor curve of historical data, on the basis of a given control target, the fitted curve can give a recommended impact factor value, such as a recommended review rate and a recommended use case density value, which is used to control the effective execution of the review and testing process of new projects, thereby ensuring the quality of software research and development. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 Design a diagram for a model-based quantitative control method for embedded software quality;

[0028] Figure 2 This is a schematic diagram of the target number of defect detections for each process segment calculated based on the organizational performance baseline;

[0029] Figure 3 A schematic diagram of the regression equation between defect density and requirement review rate is introduced for requirements;

[0030] Figure 4 A schematic diagram of the regression equation between defect density and code review rate is introduced for design implementation;

[0031] Figure 5 Schematic diagram of the regression equation of test defect density, use case density, and test progress pressure;

[0032] Figure 6 Provide a recommended review rate or use case density diagram based on the actual defect detection target value;

[0033] Figure 7 This is a schematic diagram for predicting the probability of achieving the target value of defect detection in the planning phase. DETAILED DESCRIPTION

[0034] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0035] A model-based embedded software quality quantitative control method includes the following steps:

[0036] (1) Calculate the defect detection target value for each process segment based on the organizational performance baseline;

[0037] (1.1) Complete statistical analysis of the performance baseline of defect implantation detection in each process segment;

[0038] (1.2) Calculation of estimated defect implantation values ​​for each process segment;

[0039] (1.3) Calculation of defect detection target value for each process segment.

[0040] (2) Complete the curve fitting of strain Y and influencing factor X;

[0041] (2.1) Complete the curve fitting of the number of requirements detected and the rate of requirements review;

[0042] (2.2) Complete the curve fitting of the number of code review checkouts and the code review rate;

[0043] (2.3) Complete the curve fitting of the number of system test defects detected, test case density, and test process progress pressure.

[0044] (3) Based on the project's process checkout target number, review scale, or test scale, a fitting curve is used to calculate the recommended review rate or use case density;

[0045] (3.1) Based on the target number of defects detected in the requirements process and the scale of the requirements review, the recommended requirements review rate is calculated using the fitted requirements detection number and requirements review rate curve to be used for quality control of the requirements review process;

[0046] (3.2) Based on the target number of defects detected during the design implementation process and the scale of code review, a fitted curve of the number of defects detected during the design implementation process and the code review rate is used to calculate the recommended code review rate for quality control of the code review process.

[0047] (3.3) Based on the target number of defects detected during the test process and the scale of the lines of code tested, the recommended test case density is calculated using the fitted test process detection number and test case density, and test process progress pressure rate curve for quality control of the system test process.

[0048] (4) Perform Monte Carlo simulation on the probability of achieving the target number of detections during the test process.

[0049] (4.1) During the planning phase, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detections using estimated lines of code;

[0050] (4.2) Before transitioning from development to testing, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detections after the revision using actual code lines.

[0051] Example:

[0052] The first step is to calculate the defect detection target value for each process segment based on the organizational performance baseline.

[0053] (1.1) Complete the statistical analysis of the performance baseline of defect insertion and detection in each process segment. Based on the software of the delivered projects, perform basic data statistics, including: project scale (function points), number of defects introduced in the requirements process, number of defects introduced in the AS design and implementation process, number of defects introduced in the OS design and implementation process, number of defects self-detected in the requirements process, number of defects self-detected in the AS design and implementation process, number of defects self-detected in the OS design and implementation process, and calculate the derived indicators respectively:

[0054] Demand-introduced defect density (per function point) = number of defects introduced by the demand process / project size

[0055] Defect density introduced by AS design implementation (defects / function point) = number of defects introduced by AS design implementation process / project scale

[0056] OS design and implementation defect density (number / function point) = number of defects introduced during OS design and implementation / project size

[0057] Requirements process PCE = (number of defects detected by the requirements process / number of defects introduced by the requirements process) * 100%

[0058] PCE of AS design implementation process = (number of defects detected by self-detection during AS design implementation process / number of defects introduced during AS design implementation process) * 100%

[0059] PCE of OS design and implementation process = (number of defects detected by self-detection during AS design and implementation process / number of defects introduced during OS design and implementation process) * 100%

[0060] The defect density and process PCE derived indicators are introduced into the calculated process, and the baseline values ​​are statistically analyzed using numerical statistical methods to obtain the mean and standard deviation.

[0061] (1.2) Calculation of estimated defect implantation values ​​for each process segment.

[0062] Based on the estimated function point size of the new R&D project, the estimated defect injection value of each process is obtained:

[0063] Estimated number of defects introduced by the requirements process = Demand introduction defect density (number / function point) * Estimated project size (function point)

[0064] Estimated number of defects introduced during the AS design and implementation process = Defect density introduced during the AS design and implementation process (defects / function point) * Estimated project size (function point)

[0065] Estimated number of defects introduced during OS design and implementation = Defect density introduced during OS design and implementation (defects / function point) * Estimated project size (function points)

[0066] (1.3) Calculation of defect detection target value for each process segment.

[0067] Target number of defects detected in the requirements process = Estimated number of defects introduced in the requirements process * PCE of the requirements process

[0068] Target number of defects detected during AS design implementation process = Estimated number of defects introduced during AS design implementation process * PCE of AS design implementation process

[0069] Target number of defects detected during OS design and implementation process = Estimated number of defects introduced during OS design and implementation process * PCE of OS design and implementation process

[0070] The defect detection target values ​​calculated for each process segment based on the organizational performance baseline are as follows: Figure 2 shown.

[0071] The second step is to complete the curve fitting of the strain variable Y and the influencing factor X.

[0072] (2.1) Complete the curve fitting of requirement introduction defect density and requirement review rate.

[0073] Put the historical data of requirement introduction defect density and requirement review rate into Minitab, and use the regression function of the tool to get the fitting curve formula of requirement introduction defect density Y and requirement review rate X, as shown below: Figure 3 shown.

[0074] (2.2) Complete the curve fitting of the defect density introduced by AS / OS design implementation and the code review rate. Put the historical data of the defect density introduced by AS / OS design implementation and the code review rate into the Minitab tool, and use the regression function of the tool to obtain the fitting curve formula of the defect density introduced by design implementation Y and the code review rate X, as shown in the following figure: Figure 4 shown.

[0075] (2.3) Complete the curve fitting of system test defect density, test case density, and test process progress pressure.

[0076] Put the historical data of test defect density, use case density, and test progress pressure into the Minitab tool, and use the regression function of the tool to obtain the fitting curve formula of test defect density Y, use case density X1, and test progress pressure X2, as shown below: Figure 5 shown.

[0077] The third step is to use the fitting curve to calculate the recommended review rate or use case density based on the project's process inspection target number, review scale or test scale. Figure 6 shown.

[0078] (3.1) Based on the target number of defects detected in the requirements process and the scale of the requirements review, the recommended requirements review rate is calculated using the fitted requirements detection number and requirements review rate curve, which is used for quality control of the requirements review process.

[0079] Complete the display format design of the demand review objectives, review scale, and review rate in Excel, and solidify the fitting curve formula in the cell calculation formula;

[0080] Fill in the actual scale of demand review in the corresponding column and calculate the recommended review rate based on the fitting formula.

[0081] (3.2) Based on the target number of defects detected in the design implementation process and the scale of code review, the recommended code review rate is calculated using the fitted design implementation detection number and code review rate curve for quality control of the code review process.

[0082] Complete the display format design of AS / OS code review objectives, code review scale, and code review rate in Excel, and solidify the fitting curve formula in the cell calculation formula;

[0083] Fill in the actual code review scale in the corresponding column and calculate the recommended code review rate according to the fitting formula.

[0084] (3.3) Based on the target number of defects detected during the test process and the scale of the lines of code tested, the recommended test case density is calculated using the fitted test process detection number and test case density, and test process progress pressure curve for quality control of the system test process.

[0085] Complete the display format design of test defect detection targets, test code line size, and use case density in Excel, and solidify the fitting curve formula in the cell calculation formula;

[0086] Fill the actual scale to be tested into the corresponding column and calculate the recommended test case density according to the fitting formula.

[0087] The fourth step is to perform Monte Carlo simulation on the probability of achieving the target number of detections during the test process.

[0088] (4.1) During the planning phase, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detection using estimated lines of code.

[0089] Complete the definition of the tissue average defect detection density hypothesis in Excel;

[0090] In a cell in Excel, complete the formula definition: Defect detection target number = organization average defect detection density * estimated test code lines, and define it as prediction;

[0091] Use the CRYSTAL BALL plug-in to perform Monte Carlo simulation on the predicted values;

[0092] By setting a target value (such as 16), we can get the probability (29.3%) that the simulation value is greater than the target value. Figure 7 shown.

[0093] (4.2) Before transitioning from development to testing, perform a Monte Carlo simulation of the probability of achieving the revised target number of test defect detections using actual lines of code. The implementation steps are the same as above. Set the revised target number of test defect detections (e.g., 18) and observe the probability of achieving it.

[0094] A computer storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned model-based embedded software quality quantitative control method.

[0095] A computer device includes a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the computer program, the above-mentioned model-based embedded software quality quantitative control method is implemented.

Claims

1. A model-based embedded software quality quantitative control method, characterized in that: The following steps are involved: (1) Calculate the defect detection target value for each process segment based on the organizational performance baseline; including: (1.1) Complete statistical analysis of the performance baseline of defect implantation detection in each process segment; (1.2) Calculation of estimated defect implantation values ​​for each process segment; (1.3) Calculation of defect detection target values ​​for each process segment; (2) Complete the curve fitting of the strain variable Y and the influencing factor X; including the following steps: (2.1) Complete the curve fitting of the number of requirements detected and the rate of requirements review; (2.2) Complete the curve fitting of the number of code review checkouts and the code review rate; (2.3) Complete the curve fitting of the number of system test defects detected, test case density, and test process progress pressure; (3) Based on the project's process checkout target number, review scale, or test scale, a fitting curve is used to calculate the recommended review rate or use case density; (4) Performing Monte Carlo simulation on the probability of achieving the target number of detections during the test process; including: (4.1) During the planning phase, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detections using estimated lines of code; (4.2) Before transitioning from development to testing, Monte Carlo simulation is performed on the probability of achieving the target number of test defect detections after the revision using actual code lines.

2. The method for quantitative control of embedded software quality based on a model according to claim 1, characterized in that: The step (3) comprises the following steps: (3.1) Based on the target number of defects detected in the requirements process and the scale of the requirements review, the recommended requirements review rate is calculated using the fitted requirements detection number and requirements review rate curve to be used for quality control of the requirements review process; (3.2) Based on the target number of defects detected during the design implementation process and the scale of code review, a fitted curve of the number of defects detected during the design implementation process and the code review rate is used to calculate the recommended code review rate for quality control of the code review process. (3.3) Based on the target number of defects detected during the test process and the scale of the lines of code tested, the recommended test case density is calculated using the fitted test process detection number and test case density, and test process progress pressure rate curve for quality control of the system test process.

3. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for quantitatively controlling the quality of embedded software based on a model as claimed in any one of claims 1 to 2 is implemented.

4. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the method for quantitative control of embedded software quality based on a model as described in any one of claims 1 to 2 is implemented.

Images