CN112667215A - Automatic repairing method for formalized requirement specification - Google Patents

Automatic repairing method for formalized requirement specification Download PDF

Info

Publication number
CN112667215A
CN112667215A CN202110022180.3A CN202110022180A CN112667215A CN 112667215 A CN112667215 A CN 112667215A CN 202110022180 A CN202110022180 A CN 202110022180A CN 112667215 A CN112667215 A CN 112667215A
Authority
CN
China
Prior art keywords
conflict
formula
ltl
conflicts
true
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110022180.3A
Other languages
Chinese (zh)
Other versions
CN112667215B (en
Inventor
万海
钟洪桢
罗炜麟
范科峰
赵杭
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sun Yat Sen University
Original Assignee
Sun Yat Sen University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sun Yat Sen University filed Critical Sun Yat Sen University
Publication of CN112667215A publication Critical patent/CN112667215A/en
Application granted granted Critical
Publication of CN112667215B publication Critical patent/CN112667215B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Abstract

The invention discloses an automatic repairing method for formalized requirement specifications, which comprises the following steps: s1: inputting the environmental constraint and system assertion of a requirement specification, automatically searching conflicts by using a local search algorithm, and forming a conflict set by the searched conflicts; s2: screening conflicts in the conflict set, sequencing the screened conflicts according to the occurrence probability, and taking the conflict with the highest probability; s3: negating the conflict with the highest probability obtained in the step S2 and adding the negation into the system assertion of the requirement specification; s4: returning to step S1, where the system assertion of the requirement specification is the system assertion processed in step S3, until no new conflict can be found, proceeding to step S5; s5: the environmental constraints of the demand specification and the system assertions at that time are output. The invention provides an automatic demand refinement method, which can efficiently and automatically search the conflict of demand protocols, thereby providing a direction for restoring the demand protocols.

Description

Automatic repairing method for formalized requirement specification
Technical Field
The invention relates to the field of formalized demand repair, in particular to an automatic repair method for formalized demand specifications.
Background
In a software system, software and environment always run as expected, and the realization of the software system is difficult because a situation that assertions are contradictory can be caused is inevitable, and because the software system is naturally designed in a manual design process, a phenomenon of lack of demand naturally exists, and the lack of demand is known to be one of main reasons for software failure. The requirement engineering is just a subject aiming at the problem of lack of requirements, and the main work of the requirement engineering is to perform requirement analysis according to proven effective technologies, principles and methods, so as to obtain requirement setting under the current target as much as possible, thereby increasing the robustness and completeness of software. In the formalized demand engineering, firstly, modeling is carried out on demands, and then whether the models meet the demands is analyzed. One of the analysis stages is a conflict analysis stage, the main work content of the stage is to find the problems and conflicts existing in the current system demand model, then add and modify the current demand according to the problems and conflicts, and further repair the demand model, and the conflict analysis stage mainly comprises three steps: (1) identifying conditions that cause a divergence in system assertions, i.e., a conflict; (2) assessing the severity of the conflict and the priority of the treatment; (3) and processing the conflict.
The Chinese patent with publication number CN109918049A, publication number of 2019, 06 and 21, discloses a verification method of a software development model based on formal verification, which combines the formal verification and the software development process to improve the high credibility of software and logically ensure the reliability of the developed software; the method plays a role in verifying the development logic in the software development process in the field of software error detection by formalized verification; the possibility of the program running in an abnormal state is reduced. But the invention does not give a general collision recovery method for the system.
For a better description of the requirements, the following basic concepts are explained. Linear Temporal Logic (LTL) is a fuzzy sequential logic expressed in time. A linear temporal logic formula is a logical proposition with temporal attributes. In addition to the AND (^) or (V.V.), Not (Not), implication (→), true (true), false (false) in propositional logic, LTL contains some special operators whose underlying semantic and symbolic representations are shown in Table 1. When there is a variable assignment such that the LTL formula is true, then we call this formula satisfiable, and satisfiable; when no assignment exists, the LTL formula is true, and the formula is called as mutually exclusive and is not satisfied; when any assignment can make the LTL formula true, we call this formula a perpetual formula. The satisfiability of LTL is determinable, and the primary role of the LTL solver is to determine whether an LTL formula is satisfiable. Generally speaking, given an LTL equation as an input, a solver may give a satisfiability decision for the equation, i.e., Satisfiable (SAT) and Unsatisfiable (UNSAT).
In the target-oriented demand engineering methodology, the demand specifications are derived from two parts, namely, the environment constraint (Domain property) and the system assertion (Goal) under the form of LTL (Linear Temporal logic). Environmental constraints are state descriptions of current system requirements and system assertions are assertions of system functionality. Conditions that satisfy the environmental constraint requirements but cause the system to assert a conflict are defined as a conflict. Colloquially, a conflict is a scenario that causes a software system to be in error. Environmental constraint (Domain Property): the description of the state of the current system requirement is denoted as Domain ═ dom1,dom2,...,domn}. System assertion (Goal): assertion of system function, marked as Goal ═ g1,g2,...,gn}. Conflict (Boundary Condition): a valuation scenario that satisfies the environmental constraints but causes the system to assert conflicts. Since LTL can reflect the variable in time seriesThe state changes, so it is obvious that in a system, the states of different physical quantities and their changes can be expressed using the LTL formula. Environmental constraints and system assertions are just formulas that are formally represented using LTL. We give a complete property definition of the conflict, for the system assertion set, Goal ═ g1,g2,...,gnAnd the set of environmental constraints, Domain, conflicts
Figure BDA0002888985120000021
The following properties exist:
(1) logical inconsistency:
Figure BDA0002888985120000022
is not satisfactory;
(2) minimum property:
Figure BDA0002888985120000023
is satisfactory;
(3) non-trivial:
Figure BDA0002888985120000024
is of a permanent type.
Wherein G is the conjunctive of all system assertions in the system assertion set Goal, i.e., G ═ G1∧g2∧...∧gn；G-iAsserting the conjunctive of all the other elements except the ith element in the set of good for the system, i.e. G-i＝g1∧...∧gi-1∧gi+1∧...∧gn。
TABLE 1
Figure BDA0002888985120000025
Disclosure of Invention
The invention mainly aims to provide an automatic repair method for formalized requirement specifications, which solves the problem of assertion conflict of an automatic repair system in formalized requirement engineering.
In order to solve the technical problems, the technical scheme of the invention is as follows:
an automatic repairing method for formalized requirement specification comprises the following steps:
s1: inputting the environmental constraint and system assertion of a requirement specification, automatically searching conflicts by using a local search algorithm, and forming a conflict set by the searched conflicts;
s2: screening conflicts in the conflict set, sequencing the screened conflicts according to the occurrence probability, and taking the conflict with the highest probability;
s3: negating the conflict with the highest probability obtained in the step S2 and adding the negation into the system assertion of the requirement specification;
s4: returning to step S1, where the system assertion of the requirement specification is the system assertion processed in step S3, until no new conflict can be found, proceeding to step S5;
s5: the environmental constraints of the demand specification and the system assertions at that time are output.
Preferably, the step S1 of automatically searching for a conflict by using a local search algorithm specifically includes the following steps:
s11: input the environmental constraint Domain of the requirement specification, the System asserts Goal ═ g1,g2,...,gn}；
S12: initializing the current candidate solution to Not (g)1∧g2∧...∧gn) The initial LTL formula is adopted, and the constant transform is utilized to push Not down to a variable, wherein the Not represents non-logic;
s13: randomly selecting a plurality of neighbor LTL formulas of candidate solution to form a Set;
s14: scoring each neighbor LTL formula in the Set by using a target function;
s15: adding the conflict in the Set into the result Set;
s16: updating the current candidate solution for the formula with the highest Set score;
s17: repeating steps S3-S6 until the run reaches a maximum time;
s18: and outputting the result set as a conflict set.
Preferably, the neighbor LTL formula is generated by a formula editing operation, specifically:
definition of O1Is a set of unary operators { Not, X, F, G }, λ1∈O1X, F, G are all logical symbols,
Figure BDA0002888985120000031
to represent
Figure BDA0002888985120000032
At the next moment in time that is true,
Figure BDA0002888985120000033
to represent
Figure BDA0002888985120000034
At some point in the future is true,
Figure BDA0002888985120000035
to represent
Figure BDA0002888985120000036
True at any time in the future;
definition of O2Is a binary operator set { < v, →, U, R }, λ2∈O2A, < V >, < U >, R are all logical symbols, < V > represents AND logic, < V > represents OR logic, < V > represents implied logic,
Figure BDA0002888985120000037
is shown in
Figure BDA0002888985120000038
Before being true, psi must always be true,
Figure BDA0002888985120000039
is shown in
Figure BDA00028889851200000310
Before being true, including that time instant, psi must always be true if
Figure BDA00028889851200000311
It is not possible to be true, psi remains true forever;
p is a variable, and p is a variable,
Figure BDA0002888985120000041
for the initial LTL formula, the LTL is,
Figure BDA0002888985120000042
for a new LTL formula obtained after one operation, the three formula editing operations are as follows:
a) addition of
Add a unary operator:
Figure BDA0002888985120000043
adding a binary operator and a variable:
Figure BDA0002888985120000044
or
Figure BDA0002888985120000045
b) Modifying
Modifying one variable:
Figure BDA0002888985120000046
when in use
Figure BDA0002888985120000047
When the current is over;
modify a unary operator:
Figure BDA0002888985120000048
when in use
Figure BDA0002888985120000049
When the current is over;
modify a binary operator:
Figure BDA00028889851200000410
when in use
Figure BDA00028889851200000411
When the current is over;
c) deleting
Delete a unary operator:
Figure BDA00028889851200000412
when in use
Figure BDA00028889851200000413
When the current is over;
deleting a binary operator:
Figure BDA00028889851200000414
or p' when
Figure BDA00028889851200000415
For an LTL formula
Figure BDA00028889851200000416
Any sub-formula of the LTL is subjected to the formula editing operation once to obtain a new LTL formula, and the set of the new LTL formula is formed into a neighborhood of the LTL formula.
Preferably, in step S14, each neighbor LTL formula in the Set is scored using an objective function. First, each component function of the objective function is introduced
Figure BDA00028889851200000417
Auxiliary function
Figure BDA00028889851200000418
Reintroducing the objective function
Figure BDA00028889851200000419
The method specifically comprises the following steps:
Figure BDA00028889851200000420
Figure BDA00028889851200000421
Figure BDA00028889851200000422
Figure BDA00028889851200000423
where | Goal | is the number of elements in the assertion set of the system, G-i＝g1∧...∧gi-1∧gi+1∧...∧gn。
Objective function
Figure BDA00028889851200000424
Comprises the following steps:
Figure BDA0002888985120000051
in the formula (I), the compound is shown in the specification,
Figure BDA0002888985120000052
is a formula
Figure BDA0002888985120000053
The number of variables and operators.
Preferably, when the formula
Figure BDA0002888985120000054
Is higher than 3, it is regarded as a conflict, an
Figure BDA0002888985120000055
The shorter the length of (c), the higher the score.
Preferably, in step S2, the method includes the steps of screening the conflicts in the conflict set, sorting the screened conflicts according to the occurrence probability, and selecting the conflict with the highest probability, and specifically includes the following steps:
s21: inputting a conflict set sigma and outputting a general conflict set with the highest probability;
s22: traversing each LTL formula in Σ
Figure BDA0002888985120000056
If found using LTL solver
Figure BDA0002888985120000057
Figure BDA00028889851200000511
The presentation logic is to push out a symbol,
Figure BDA0002888985120000058
logically deducing psi, then deleting
Figure BDA0002888985120000059
Finally forming a general conflict set by the residual LTL formulas of the sigma;
s23: calculating the probability of each conflict in the general conflict set by using a model counting method;
s24: and selecting the general conflict with the highest probability and outputting.
Preferably, in step S23, a model counting method is used to calculate the probability of each collision in the general collision set, specifically:
respectively calculating the number # a of models which are asserted by an environment constraint inverted V system and the number # b of models which are asserted by the environment constraint inverted V system within the limited model length by using a model counting method,
Figure BDA00028889851200000510
compared with the prior art, the technical scheme of the invention has the beneficial effects that:
the invention provides an automatic demand refinement method, which can efficiently and automatically search and describe the conflict of demand protocols, thereby providing a direction for demand restoration. On the aspect of the problem of automatic search conflict, the local search algorithm designed by the invention greatly improves the search performance.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention.
Detailed Description
The drawings are for illustrative purposes only and are not to be construed as limiting the patent;
for the purpose of better illustrating the embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product;
it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.
Example 1
The embodiment provides an automatic repairing method for formalized requirement specifications, as shown in fig. 1, including the following steps:
s1: inputting the environmental constraint and system assertion of a requirement specification, automatically searching conflicts by using a local search algorithm, and forming a conflict set by the searched conflicts;
s2: screening conflicts in the conflict set, sequencing the screened conflicts according to the occurrence probability, and taking the conflict with the highest probability;
s3: negating the conflict with the highest probability obtained in the step S2 and adding the negation into the system assertion of the requirement specification;
s4: returning to step S1, where the system assertion of the requirement specification is the system assertion processed in step S3, until no new conflict can be found, proceeding to step S5;
s5: the environmental constraints of the demand specification and the system assertions at that time are output.
In step S1, the conflict is automatically found by using a local search algorithm, which specifically includes the following steps:
s11: the environment constraint Domain of the input requirement, the system declares the good ═ g1,g2,...,gn}；
S12: initializing the current candidate solution to Not (g)1∧g2∧...∧gn) Is the initial LTL formula and utilizes constantsThe transformation pushes Not down to the variable, Not representing Not logic;
s13: randomly selecting a plurality of neighbor LTL formulas of candidate solution to form a Set;
s14: scoring each neighbor LTL formula in the Set by using a target function;
s15: adding the conflict in the Set into the result Set;
s16: updating the current candidate solution for the formula with the highest Set score;
s17: repeating steps S3-S6 until the run reaches a maximum time;
s18: and outputting the result set as a conflict set.
The neighbor LTL formula is generated by formula editing operation, and specifically comprises the following steps:
definition of O1Is a set of unary operators { not, X, F, G }, λ1∈O1X, F, G are all logical symbols,
Figure BDA0002888985120000061
to represent
Figure BDA0002888985120000062
At the next moment in time that is true,
Figure BDA0002888985120000063
to represent
Figure BDA0002888985120000064
At some point in the future is true,
Figure BDA0002888985120000065
to represent
Figure BDA0002888985120000066
True at any time in the future;
definition of O2Is a binary operator set { < v, →, U, R }, λ2∈O2A, < V >, < U >, R are all logical symbols, < V > represents AND logic, < V > represents OR logic, < V > represents implied logic,
Figure BDA0002888985120000067
is shown in
Figure BDA0002888985120000068
Before being true, psi must always be true,
Figure BDA0002888985120000069
is shown in
Figure BDA00028889851200000610
Before being true, including that time instant, psi must always be true if
Figure BDA00028889851200000611
It is not possible to be true, psi remains true forever;
p is a variable, and p is a variable,
Figure BDA00028889851200000612
for the initial LTL formula, the LTL is,
Figure BDA00028889851200000613
for a new LTL formula obtained after one operation, the three formula editing operations are as follows:
a) addition of
Add a unary operator:
Figure BDA0002888985120000071
adding a binary operator and a variable:
Figure BDA0002888985120000072
or
Figure BDA0002888985120000073
b) Modifying
Modifying one variable:
Figure BDA0002888985120000074
when in use
Figure BDA0002888985120000075
When the current is over;
modify a unary operator:
Figure BDA0002888985120000076
when in use
Figure BDA0002888985120000077
When the current is over;
modify a binary operator:
Figure BDA0002888985120000078
when in use
Figure BDA0002888985120000079
When the current is over;
c) deleting
Delete a unary operator:
Figure BDA00028889851200000710
when in use
Figure BDA00028889851200000711
When the current is over;
deleting a binary operator:
Figure BDA00028889851200000712
or p' when
Figure BDA00028889851200000713
For an LTL formula
Figure BDA00028889851200000714
Any sub-formula of the LTL is subjected to the formula editing operation once to obtain a new LTL formula, and the set of the new LTL formula is formed into a neighborhood of the LTL formula.
In step S14, each neighbor LTL formula in the Set is scored using an objective function. First, each component function of the objective function is introduced
Figure BDA00028889851200000715
Auxiliary function
Figure BDA00028889851200000716
Reintroducing the objective function
Figure BDA00028889851200000717
The method specifically comprises the following steps:
Figure BDA00028889851200000718
Figure BDA00028889851200000719
Figure BDA00028889851200000720
Figure BDA00028889851200000721
where | Goal | is the number of elements in the assertion set of the system, G-i＝g1∧...∧gi-1∧gi+1∧...∧gn. Objective function
Figure BDA00028889851200000722
Comprises the following steps:
Figure BDA00028889851200000723
in the formula (I), the compound is shown in the specification,
Figure BDA00028889851200000724
is a formula
Figure BDA00028889851200000725
The number of variables and operators.
When the formula
Figure BDA00028889851200000726
Is higher than 3, it is regarded as a conflict, an
Figure BDA00028889851200000727
The shorter the length of (c), the higher the score.
The prior art has two kinds of conflict of automatic search, one is to adopt a table decomposition method, and the other is to adopt a genetic algorithm. The table decomposition method belongs to a construction method, and expands the LTL formula semantics into a syntax tree with a back edge, thereby constructing a conflict. The genetic algorithm belongs to a search method, and initializes a batch of LTL formulas, and simulates chromosome behaviors through the formulas to carry out mutation and cross exchange so as to obtain conflicts. The invention adopts a local search algorithm in the automatic search conflict, also belongs to a search method, and is different from a genetic algorithm in that the direction of formula change is appointed by a formula neighborhood, a target function is appointed to lead the search to advance towards the conflict, and the conflict of the local search is appointed, so that the conflict can be found with less iteration times.
In step S2, the method includes the steps of screening conflicts in the conflict set, sorting the screened conflicts according to the occurrence probability, and selecting the conflict with the highest probability, which specifically includes the following steps:
s21: inputting a conflict set sigma and outputting a general conflict set with the highest probability;
s22: traversing each LTL formula in Σ
Figure BDA0002888985120000081
If found using LTL solver
Figure BDA0002888985120000082
Figure BDA0002888985120000086
The presentation logic is to push out a symbol,
Figure BDA0002888985120000083
logically deducing psi, then deleting
Figure BDA0002888985120000084
Finally forming a general conflict set by the residual LTL formulas of the sigma;
s23: calculating the probability of each conflict in the general conflict set by using a model counting method;
s24: and selecting the general conflict with the highest probability and outputting.
In step S23, a model counting method is used to calculate the probability of each collision in the general collision set, specifically:
respectively calculating the number # a of models which are asserted by an environment constraint inverted V system and the number # b of models which are asserted by the environment constraint inverted V system within the limited model length by using a model counting method,
Figure BDA0002888985120000085
in a particular embodiment, a water pump system in a mine tunnel is considered, the system being aimed at preventing flood bursts in the mine tunnel. But when the system detects that the water level is too high, the water pump is started to pump water. But when the system detects that gas exists in the mine hole, the water pump cannot be started. Now, the water level is represented by a Boolean variable h, p represents that the water pump is started, and m represents that gas exists in the mine hole. Then the environmental constraints and system assertions are as follows:
and (3) environmental constraint: when the water pump is continuously started for two times, the water level is changed to a low water level at the third time
The LTL formula describes: g ((p. lamda. Xp) → X (X Not h))
The system asserts 1: when the water level is too high, the system starts the water pump
The LTL formula describes: g (h → Xp)
The system asserts 2: when gas exists in the mine hole, the system can not start the water pump
The LTL formula describes: g (m → X Not h)
The steps of the process using this embodiment are as follows:
the method comprises the following steps: and inputting an environment constraint and a system assertion set after formalized requirements, and positioning conflicts by using a local search algorithm.
1. Initializing the current candidate formula to be Not (G (h → Xp) ^ G (m → X Not h));
2. the current candidate formula is rewritten as F (h ^ X (Notp)) V-V (m ^ X (p)) by removing → and pushing Not down on the variable with the constant transform;
3. obtaining a plurality of neighbors of the current candidate formula by using formula editing operation, wherein one neighbor is as follows:
F(h∧X(Not p))∨(m∧X(p))
deleting F of the current candidate formula to obtain;
calculating the score of the neighbor formula, F (h Λ x (not p)) (m Λ x (p))) score of 1+1+1+0.09 ═ 3.09, i.e. it is a conflict, added to the result set;
4. and modifying the current candidate formula into the formula with the highest score in the neighborhood. Assuming that the formula with the highest score among the currently selected neighbors is F (h ^ X (Notp)) (m ^ X (p)), then this formula is called the current candidate formula for the next iteration.
5. And repeating the steps 3-4, and finally outputting a result set as a conflict set.
And step two, screening and sorting the conflict set.
Suppose step one outputs a set of conflicts as
{F(h∧m),h∧m,F(h∧Notm∧p∧X((Not h∧Not p)∨(h∧(m∨Not p))))}
1. Only is provided with
Figure BDA0002888985120000092
Thus deleting h ^ m
2. Obtain the general conflict set { F (h ^ m), F (h ^ Not m ^ p ^ X ((Not h ^ Not p) ((h ^ m) (Not))))) }
3. The probability of collision occurrence within the length of the model 1000 is calculated by using a model counting method as shown in table 2:
TABLE 2
Figure BDA0002888985120000091
4. The general conflict F with the highest output probability (h ^ Not m ^ p ^ X ((Not h ^ Not p) V (h ^ Not p)))) is thus obtained.
Step three, selecting the general conflict with the highest probability, negating the general conflict, and adding the general conflict into a system assertion set;
not (F (h ^ Not m ^ p ^ x ((Not h ^ Notp) V-shaped (h ^ (m ^ Notp)))) is added into the original system assertion set as a newly added system assertion.
Step four: and repeating the first step to the third step until no conflict is found.
The same or similar reference numerals correspond to the same or similar parts;
the terms describing positional relationships in the drawings are for illustrative purposes only and are not to be construed as limiting the patent;
it should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.

Claims (7)

1. An automatic repairing method for formalized requirement specification is characterized by comprising the following steps:
s1: inputting the environmental constraint and system assertion of a requirement specification, automatically searching conflicts by using a local search algorithm, and forming a conflict set by the searched conflicts;
s2: screening conflicts in the conflict set, sequencing the screened conflicts according to the occurrence probability, and taking the conflict with the highest probability;
s3: negating the conflict with the highest probability obtained in the step S2 and adding the negation into the system assertion of the requirement specification;
s4: returning to step S1, where the system assertion of the requirement specification is the system assertion processed in step S3, until no new conflict can be found, proceeding to step S5;
s5: the environmental constraints of the demand specification and the system assertions at that time are output.
2. The automated formal requirement specification repair method according to claim 1, wherein the step S1 of automatically searching for conflicts by using a local search algorithm includes the following steps:
s11: input the environmental constraint Domain of the requirement specification, the System asserts Goal ═ g1,g2,...,gn}；
S12: initializing the current candidate solution to Not (g)1∧g2∧...∧gn) The initial LTL formula is adopted, and the constant transform is utilized to push Not down to a variable, wherein the Not represents non-logic;
s13: randomly selecting a plurality of neighbor LTL formulas of candidate solution to form a Set;
s14: scoring each neighbor LTL formula in the Set by using a target function;
s15: adding the conflict in the Set into the result Set;
s16: updating the current candidate solution for the formula with the highest Set score;
s17: repeating steps S3-S6 until the run reaches a maximum time;
s18: and outputting the result set as a conflict set.
3. The method for automated repair of formal requirement conventions according to claim 2, wherein the neighbor LTL formula is generated by a formula editing operation, specifically:
definition of O1Is a set of unary operators { Not, X, F, G }, λ1∈O1X, F, G are all logical symbols,
Figure FDA0002888985110000011
to represent
Figure FDA0002888985110000012
At the next moment in time that is true,
Figure FDA0002888985110000013
to represent
Figure FDA0002888985110000014
At some point in the future is true,
Figure FDA0002888985110000015
to represent
Figure FDA0002888985110000016
True at any time in the future;
definition of O2Is a binary operator set { < v, →, U, R }, λ2∈O2A, < V >, < U >, R are all logical symbols, < V > represents AND logic, < V > represents OR logic, < V > represents implied logic,
Figure FDA0002888985110000017
is shown in
Figure FDA0002888985110000018
Before being true, psi must always be true,
Figure FDA0002888985110000019
is shown in
Figure FDA00028889851100000110
Before being true, including that time instant, psi must always be true if
Figure FDA00028889851100000111
It is not possible to be true, psi remains true forever;
p is a variable, and p is a variable,
Figure FDA0002888985110000021
for the initial LTL formula, the LTL is,
Figure FDA0002888985110000022
for a new LTL formula obtained after one operation, the three formula editing operations are as follows:
a) addition of
Add a unary operator:
Figure FDA0002888985110000023
adding a binary operator and a variable:
Figure FDA0002888985110000024
or
Figure FDA0002888985110000025
b) Modifying
Modifying one variable:
Figure FDA0002888985110000026
when in use
Figure FDA0002888985110000027
When the current is over;
modify a unary operator:
Figure FDA0002888985110000028
when in use
Figure FDA0002888985110000029
When the current is over;
modify a binary operator:
Figure FDA00028889851100000210
when in use
Figure FDA00028889851100000211
When the current is over;
c) deleting
Deleting a unary operationSymbol:
Figure FDA00028889851100000212
when in use
Figure FDA00028889851100000213
When the current is over;
deleting a binary operator:
Figure FDA00028889851100000214
or p' when
Figure FDA00028889851100000215
For an LTL formula
Figure FDA00028889851100000216
Any sub-formula of the LTL is subjected to the formula editing operation once to obtain a new LTL formula, and the set of the new LTL formula is formed into a neighborhood of the LTL formula.
4. The method for automated repair of formal requirement conventions according to claim 3, wherein in step S14, the objective function is used to score each neighbor LTL formula in the Set, and each score function of the objective function is introduced first
Figure FDA00028889851100000217
Auxiliary function
Figure FDA00028889851100000218
Reintroducing the objective function
Figure FDA00028889851100000219
The method specifically comprises the following steps:
Figure FDA00028889851100000220
Figure FDA00028889851100000221
Figure FDA00028889851100000222
Figure FDA00028889851100000223
where | Goal | is the number of elements in the assertion set of the system, G-i＝g1∧...∧gi-1∧gi+1∧...∧gn；
Objective function
Figure FDA00028889851100000224
Comprises the following steps:
Figure FDA0002888985110000031
in the formula (I), the compound is shown in the specification,
Figure FDA0002888985110000032
is a formula
Figure FDA0002888985110000033
The number of variables and operators.
5. The formalized requirement specification automated repair method according to claim 4, wherein when the formula is
Figure FDA0002888985110000034
Is higher than 3, it is regarded as a conflict, an
Figure FDA0002888985110000035
The shorter the length of (c), the higher the score.
6. The method for automatically repairing formalized requirement stipulations according to claim 5, wherein the step S2 is to screen conflicts in the conflict set, sort the screened conflicts according to the probability of occurrence, and select the conflict with the highest probability, and specifically comprises the following steps:
s21: inputting a conflict set sigma and outputting a general conflict set with the highest probability;
s22: traversing each LTL formula in Σ
Figure FDA0002888985110000036
If phi e sigma can be found using the LTL solver,
Figure FDA0002888985110000037
Figure FDA0002888985110000038
the presentation logic is to push out a symbol,
Figure FDA0002888985110000039
logically deducing psi, then deleting
Figure FDA00028889851100000310
Finally forming a general conflict set by the residual LTL formulas of the sigma;
s23: calculating the probability of each conflict in the general conflict set by using a model counting method;
s24: and selecting the general conflict with the highest probability and outputting.
7. The automated formal requirement specification repair method according to claim 6, wherein in step S23, a model counting method is used to calculate the probability of each conflict in the general conflict set, specifically:
respectively calculating the length of limited model and the environmental constraint A within the limited model by using a model counting methodThe number of models # a of system assertion Λ conflict and the number of models # b of environment constraint Λ system assertion,
Figure FDA00028889851100000311
CN202110022180.3A 2020-12-11 2021-01-08 Automatic repairing method for formalized requirement specification Active CN112667215B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202011453300 2020-12-11
CN2020114533007 2020-12-11

Publications (2)

Publication Number Publication Date
CN112667215A true CN112667215A (en) 2021-04-16
CN112667215B CN112667215B (en) 2022-02-25

Family

ID=75413747

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110022180.3A Active CN112667215B (en) 2020-12-11 2021-01-08 Automatic repairing method for formalized requirement specification

Country Status (1)

Country Link
CN (1) CN112667215B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114356294A (en) * 2021-12-21 2022-04-15 华东师范大学 Instance generation method and system based on FQLTL language

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030225552A1 (en) * 2002-05-30 2003-12-04 Nec Corporation Efficient approaches for bounded model checking
US20080271000A1 (en) * 2007-04-25 2008-10-30 Microsoft Corporation Predicting Conflicts in a Pervasive System
US20160371167A1 (en) * 2015-06-17 2016-12-22 General Electric Company Scalable methods for analyzing formalized requirements and localizing errors
CN109388377A (en) * 2018-10-08 2019-02-26 广西民族大学 A kind of software security requirement acquisition method based on Formal Modeling
CN109918049A (en) * 2019-01-12 2019-06-21 西北工业大学 The verification method of software development model based on Formal Verification
US20190317760A1 (en) * 2018-04-17 2019-10-17 The Regents Of The University Of Michigan Interactive And Dynamic Search Based Approach To Software Refactoring Recommendations
CN110532167A (en) * 2019-07-05 2019-12-03 华东师范大学 A kind of state machine model timing property verification method based on model conversion

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030225552A1 (en) * 2002-05-30 2003-12-04 Nec Corporation Efficient approaches for bounded model checking
US20080271000A1 (en) * 2007-04-25 2008-10-30 Microsoft Corporation Predicting Conflicts in a Pervasive System
US20160371167A1 (en) * 2015-06-17 2016-12-22 General Electric Company Scalable methods for analyzing formalized requirements and localizing errors
US20190317760A1 (en) * 2018-04-17 2019-10-17 The Regents Of The University Of Michigan Interactive And Dynamic Search Based Approach To Software Refactoring Recommendations
CN109388377A (en) * 2018-10-08 2019-02-26 广西民族大学 A kind of software security requirement acquisition method based on Formal Modeling
CN109918049A (en) * 2019-01-12 2019-06-21 西北工业大学 The verification method of software development model based on Formal Verification
CN110532167A (en) * 2019-07-05 2019-12-03 华东师范大学 A kind of state machine model timing property verification method based on model conversion

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
RENZO DEGIOVANNI ET AL.: "A Genetic Algorithm for Goal-Conflict Identification", 《HTTPS://DOI.ORG/10.1145/3238147.3238220》 *
RENZO DEGIOVANNI ET AL.: "Goal-Conflict Detection based on Temporal Satisfiability Checking", 《2016 31ST IEEE/ACM INTERNATIONAL CONFERENCE ON AUTOMATED SOFTWARE ENGINEERING (ASE)》 *
RENZO DEGIOVANNI ET AL.: "Goal-Conflict Likelihood Assessment based on Model Counting", 《HTTPS://DOI.ORG/10.1145/3180155.3180261》 *
文志诚: "面向对象软件的形式验证技术", 《中国优秀博硕士学位论文全文数据库(博士)信息科技辑》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114356294A (en) * 2021-12-21 2022-04-15 华东师范大学 Instance generation method and system based on FQLTL language

Also Published As

Publication number Publication date
CN112667215B (en) 2022-02-25

Similar Documents

Publication Publication Date Title
CN109918127B (en) Defect error correction method based on code modification mode difference
CN109991951B (en) Multi-source fault detection and diagnosis method and device
CN117951701A (en) Method for determining flaws and vulnerabilities in software code
Tang et al. Adversarial attack on hierarchical graph pooling neural networks
CN102598001A (en) Techniques for analysis of logic designs with transient logic
CN112667215B (en) Automatic repairing method for formalized requirement specification
Gurumurthy et al. Fair simulation minimization
US10540468B1 (en) Verification complexity reduction via range-preserving input-to-constant conversion
Kim et al. Application of natural language processing (NLP) and text-mining of big-data to engineering-procurement-construction (EPC) bid and contract documents
Tran et al. Parallel symmetric class expression learning
CN114579430A (en) Test case recommendation system and method based on pre-training language model
CN114329482A (en) C/C + + vulnerability based on sequencing and inter-patch link recovery system and method thereof
CN116305158A (en) Vulnerability identification method based on slice code dependency graph semantic learning
Niloofar et al. Fusion of data and expert knowledge for fault tree reliability analysis of cyber-physical systems
CN113094863A (en) Civil aircraft system operation reliability assessment method considering failure propagation
CN114579761A (en) Information security knowledge entity relation connection prediction method, system and medium
CN112948193B (en) FPGA comprehensive tool defect detection method based on difference test
CN114564202A (en) Symbol model detection method and system based on SAT solver and application thereof
Luo et al. Compact Abstract Graphs for Detecting Code Vulnerability with GNN Models
Kim et al. Improving the Robustness of the Bug Triage Model through Adversarial Training
Hao et al. A novel vulnerability severity assessment method for source code based on a graph neural network
Viet et al. Transfer learning for predicting software faults
Bogaerts et al. SAT-to-SAT in QBFEval 2016.
de la Higuera et al. Inference of ω-languages from prefixes
US20240054369A1 (en) Ai-based selection using cascaded model explanations

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant