CN105653793A - Random verification method and apparatus - Google Patents
Random verification method and apparatus Download PDFInfo
- Publication number
- CN105653793A CN105653793A CN201511019072.1A CN201511019072A CN105653793A CN 105653793 A CN105653793 A CN 105653793A CN 201511019072 A CN201511019072 A CN 201511019072A CN 105653793 A CN105653793 A CN 105653793A
- Authority
- CN
- China
- Prior art keywords
- vector
- random constraints
- random
- coverage rate
- pond
- 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
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/32—Circuit design at the digital level
- G06F30/33—Design verification, e.g. functional simulation or model checking
Abstract
Embodiments of the invention provide a random verification method and apparatus, and relate to the field of chip design. The method comprises the steps of determining at least four random constraint vectors contained in a random constraint pool, wherein the random constraint pool contains at least one random constraint parameter; performing simulation processing on the random constraint vectors in the random constraint pool to obtain an initial coverage rate; when the initial coverage rate does not reach a preset target coverage rate, circularly performing differential evolution processing on the random constraint pool, obtaining simulation condition parameters, and determining whether the simulation condition parameters meet a termination requirement or not until the simulation condition parameters meet the termination requirement, wherein the termination requirement is preset requirement information used for terminating the operation of performing differential evolution processing on the random constraint pool; and according to the random constraint vectors in the random constraint pool subjected to differential evolution processing, performing random verification. The method is suitable for a scene of performing random verification by using a plurality of random constraints.
Description
Technical field
The present invention relates to chip design field, particularly relate to method and the device of a kind of accidental validation.
Background technology
Along with the development of science and technology, the function of the various network equipments becomes increasingly complex, and therefore the complexity of the chip in the network equipment improves constantly. Accordingly, the complexity of checking work chip tested also improves constantly. For the chip that complexity is high, it is necessary to adopt the mode gathering checking to be verified.
In traditional accidental validation mode, one time accidental validation adopts one group of random constraints, with assessment verification the verifying results after terminating. Can there is this problem how determining that checking scale is the number determining arbitrary excitation in this mode. If the number of the arbitrary excitation of an accidental validation is very little, that is it is possible that the problem that all cannot cover of the scene that much should be able to cover. If the number of the arbitrary excitation of an accidental validation is too many, due to be same group of random constraints, it may appear that the problem that the multiformity of the scene of generation exponentially declines over time. Adding " number of suitable arbitrary excitation " is can be changing along with current checking degree and specific random constraints change, it is impossible to obtain precise figure by analyzing with previous experiences. Therefore, currently general mode is when sacrificing verification efficiency, ensures verification the verifying results as much as possible, is and selects accidental validation excitation scale big as far as possible.
Summary of the invention
Embodiments of the invention provide method and the device of a kind of accidental validation, in order to, while ensureing verification the verifying results, to improve verification efficiency.
For reaching above-mentioned purpose, embodiments of the invention adopt the following technical scheme that
A kind of method embodiments providing accidental validation, comprises determining that out at least four random constraints vector comprised in random constraints pond; Described random constraints vector includes at least one random constraints parameter; Respectively the random constraints vector in described random constraints pond is carried out simulation process, obtain initial coverage rate;When described initial coverage rate is not up to goal-selling coverage rate, described random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, it is determined that whether described simulated conditions parameter meets is terminated requirement, until simulated conditions parameter meets terminates requirement; Described end requires to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process; Random constraints vector in described random constraints pond after processing according to differential evolution, carries out accidental validation.
Alternatively, described described random constraints pond is carried out differential evolution process include: select in described random constraints pond first random constraints vector and the second random constraints vector, calculate choose described first random constraints vector and the second random constraints vector difference vector; It is weighted processing to described difference vector according to weighter factor, and according to default variation rule, utilizes the described difference vector after weighting and the 3rd random constraints vector, it is determined that variation vector; Described first random constraints vector and the second random constraints vector are any two random constraints vectors in described random constraints pond; Described 3rd random constraints vector is other any one the random constraints vectors except described first random constraints vector and the second random constraints vector in described random constraints pond; By described variation vector and the 4th random constraints vector, according to crossover probability CR, carry out mixed processing, obtain trial vector; Described 4th random constraints vector is other any one random constraints vector except described first random constraints vector, outside the second random constraints vector and the 3rd random constraints vector in described random constraints pond; Simulation process is carried out, the coverage rate after acquisition process according to described trial vector; Determine that whether the coverage rate after described process is more than described initial coverage rate; If the coverage rate after described process is more than described initial coverage rate, then according to described trial vector, update described random constraints pond.
Alternatively, described end requires to include: coverage rate reaches described goal-selling coverage rate; Described random constraints pond is carried out differential evolution process by described circulation, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets to terminate, require to include until simulated conditions parameter meets end: described random constraints pond is carried out differential evolution process by circulation, and obtain the coverage rate after process, determine whether the coverage rate after described process reaches goal-selling coverage rate, until the coverage rate after processing reaches described goal-selling coverage rate.
Alternatively, described random constraints pond is carried out differential evolution process by described circulation, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets to terminate, require to include until simulated conditions parameter meets end: described random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets and terminate requirement, and simulated conditions parameter be unsatisfactory for end require time, update weighter factor and CR, until simulated conditions parameter meets terminates requirement.
Alternatively, the initial value of described weighter factor is 0.7; The initial value of described CR is 0.8.
Further, embodiments provide the device of a kind of accidental validation, comprise determining that unit, for determining at least four random constraints vector comprised in random constraints pond; Described random constraints vector includes at least one random constraints parameter; Processing unit, for respectively the random constraints vector in described random constraints pond being carried out simulation process, obtains initial coverage rate; Described processing unit, it is additionally operable to when described initial coverage rate is not up to goal-selling coverage rate, described random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets and terminate requirement, until simulated conditions parameter meets terminates requirement; Described end requires to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process; Authentication unit, for the random constraints vector in the described random constraints pond after processing according to differential evolution, carries out accidental validation.
Alternatively, described processing unit, specifically for selecting the first random constraints vector and the second random constraints vector in described random constraints pond, calculate the difference vector of described first random constraints vector and the second random constraints vector chosen;It is weighted processing to described difference vector according to weighter factor, and according to default variation rule, utilizes the described difference vector after weighting and the 3rd random constraints vector, it is determined that variation vector; Described first random constraints vector and the second random constraints vector are any two random constraints vectors in described random constraints pond; Described 3rd random constraints vector is other any one the random constraints vectors except described first random constraints vector and the second random constraints vector in described random constraints pond; By described variation vector and the 4th random constraints vector, according to crossover probability CR, carry out mixed processing, obtain trial vector; Described 4th random constraints vector is other any one random constraints vector except described first random constraints vector, outside the second random constraints vector and the 3rd random constraints vector in described random constraints pond; Simulation process is carried out, the coverage rate after acquisition process according to described trial vector; Determine that whether the coverage rate after described process is more than described initial coverage rate; If the coverage rate after described process is more than described initial coverage rate, then according to described trial vector, update described random constraints pond.
Alternatively, described end requires to include: coverage rate reaches described goal-selling coverage rate; Described processing unit, specifically for circulation, described random constraints pond is carried out differential evolution process, and obtain the coverage rate after process, it is determined that whether the coverage rate after described process reaches goal-selling coverage rate, until the coverage rate after processing reaches described goal-selling coverage rate.
Alternatively, described processing unit, specifically for circulation, described random constraints pond is carried out differential evolution process, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets and terminate requirement, and simulated conditions parameter be unsatisfactory for end require time, update weighter factor and CR, until simulated conditions parameter meets terminates requirement.
Alternatively, the initial value of described weighter factor is 0.7; The initial value of described CR is 0.8.
Embodiments provide method and the device of a kind of accidental validation, comprise determining that out at least four random constraints vector comprised in random constraints pond; Respectively the random constraints vector in random constraints pond is carried out simulation process, obtain initial coverage rate; When initial coverage rate does not arrive goal-selling coverage rate, random constraints pond is carried out differential evolution process by circulation, and obtains simulated conditions parameter, it is determined that whether simulated conditions parameter meets is terminated, until simulated conditions parameter meets terminates requirement; Terminate to require to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process; Random constraints vector in described random constraints pond after processing according to differential evolution, carries out accidental validation. So, in the accidental validation that random constraints vector is more complicated, it is achieved that the self-adaptative adjustment to random constraints vector. Under the premise that need not increase labor workload, can emulate with different good random constraints vectors all the time, solve the waste of time that long unsuitable random constraints vector brings and resource, thus drastically increasing the efficiency of accidental validation. And then achieve while ensureing verification the verifying results, improve the purpose of verification efficiency.
Accompanying drawing explanation
In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, the accompanying drawing used required in embodiment or description of the prior art will be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the premise not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
The schematic flow sheet of the method for a kind of accidental validation that Fig. 1 provides for the embodiment of the present invention;
The schematic flow sheet of the method for the another kind of accidental validation that Fig. 2 provides for the embodiment of the present invention;
The schematic flow sheet of the method for the another kind of accidental validation that Fig. 3 provides for the embodiment of the present invention;
The structural representation of the device of a kind of accidental validation that Fig. 4 provides for the embodiment of the present invention.
Detailed description of the invention
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments. Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under not making creative work premise, broadly fall into the scope of protection of the invention.
A kind of method embodiments providing accidental validation, as it is shown in figure 1, include:
Step 101, determine random constraints pond comprises at least four random constraints vector.
Wherein, random constraints vector includes at least one random constraints parameter.
Concrete, the random constraints vector comprised in random constraints pond is more many, then the follow-up random constraints multiformity that can generate is more high, but convergence also can be slow, it is therefore desirable to determine the number of the random constraints vector comprised in random constraints pond according to the demand that realizes of user. Value after the number of the random constraints vector comprised in determining random constraints pond, to random constraints vector concrete in random constraints pond, it is possible to by the program stochastic generation write. Manual analysis can also be passed through, manually generate the random constraints value that diversity factor is relatively larger.
It should be noted that owing to, in follow-up process, once complete differential evolution processes needs four different random constraints vectors, therefore need to comprise at least four random constraints vector in random constraints pond.
Step 102, respectively the random constraints vector in random constraints pond is carried out simulation process, obtain initial coverage rate.
Concrete, after the device of accidental validation multiple random constraints vectors in determining random constraints pond, it is possible to all random constraints vector in random constraints pond is carried out respectively the simulation process of T time, such that it is able to obtain initial coverage rate.
It should be noted that after how the device of accidental validation to carry out simulation process, obtaining initial coverage rate is prior art, and the present invention does not repeat them here.
Pre-set according to the actual requirements it should be noted that T time is user.
Step 103, when initial coverage rate does not arrive goal-selling coverage rate, random constraints pond is carried out differential evolution process by circulation, and obtains simulated conditions parameter, it is determined that whether simulated conditions parameter meets is terminated requirement, until simulated conditions parameter meets terminates requirement.
Wherein, terminate to require to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process.
Concrete, the device of accidental validation is after obtaining initial coverage rate, it is possible to initial coverage rate and goal-selling coverage rate are compared, it is determined that whether initial coverage rate reaches goal-selling coverage rate. When determining that initial coverage rate is not reaching to goal-selling coverage rate, illustrate that the random constraints vector in random constraints pond is not preferably random constraints variable, now the random vector in random constraints pond can be carried out differential evolution process. When the random constraints vector in random constraints pond being carried out differential evolution and processing, can carry out repeatedly, can only carry out once, now user can pre-set end requirement, processes such that it is able to the device controlling accidental validation needs that the random constraints vector in random constraints pond carries out differential evolution several times. Therefore, the device of accidental validation is when determining that initial coverage rate is not reaching to goal-selling coverage rate, random constraints pond can be carried out differential evolution process, and obtain simulated conditions parameter, determine that whether simulated conditions parameter meets and terminate requirement, simulated conditions parameter meet end require time, illustrate that the device of accidental validation is without carrying out differential evolution process to the random constraints vector in random constraints pond, at this point it is possible to carry out next step. And simulated conditions parameter be unsatisfactory for end require time, illustrate the device of accidental validation the random constraints vector in random constraints pond has been carried out differential evolution process after, also the random constraints vector in random constraints pond need to carried out differential evolution process. Now, the device of accidental validation needs again random constraints pond to be carried out differential evolution process, and after this has carried out differential evolution process, then reacquire simulated conditions parameter, again determine whether the simulated conditions parameter of reacquisition meets and terminate requirement. And reacquire simulated conditions parameter be unsatisfactory for end require time, again random constraints pond is being carried out differential evolution process, until obtain simulated conditions parameter meet terminate requirement.
It should be noted that simulated conditions parameter is according to the parameter terminating to require acquisition.Such as, when terminating to require to reach Preset Time for simulation time, simulated conditions parameter is the simulation time having be carried out. Terminate to require that the number of times for random constraints pond carries out differential evolution process reaches preset times, then now simulated conditions parameter is the number of times that random constraints pond is carried out differential evolution process by the device of accidental validation. The present invention is without limitation.
Further, differential evolution algorithm is a kind of emerging evolutionary computation technique, being mainly used in solving the Global Optimal Problem of continuous variable, its groundwork step includes variation (Mutation), intersects (Crossover), selects (Selection) three kinds of operations. The basic thought of algorithm is from a certain initial population randomly generated, utilize two the individual difference vectors randomly selected from population as the 3rd individual change at random source, producing variation individuality by suing for peace with the 3rd individuality according to certain rule after difference vector weighting, this operation is called variation. Then, the target individual that variation individuality predetermines with certain carries out parameter and mixes, and generates test individuality, and this process is referred to as to intersect. If the fitness value of test individuality is better than the fitness value of target individual, then in the next generation, test individuality replaces target individual, and otherwise target individual still preserves, and this operation is called selection. In the evolutionary process of every generation, as target individual once, algorithm, by constantly iterative computation, retains defect individual to each voxel vector, eliminates worst individual, and guiding search process is approached to globally optimal solution.
Based on this, include as in figure 2 it is shown, random constraints pond to be carried out differential evolution process:
Random constraints pond selects the first random constraints vector and the second random constraints vector, calculates the difference vector of the first random constraints vector and the second random constraints vector chosen; It is weighted processing to difference vector according to weighter factor, and according to preset rules, utilizes the difference vector after weighting and the 3rd random constraints vector, it is determined that variation vector.
Wherein, the first random constraints vector and the second random constraints vector are any two random constraints vectors in random constraints pond; 3rd random constraints vector is other any one the random constraints vectors except the first random constraints vector and the second random constraints vector in random constraints pond.
Concrete, said process is the mutation operation that differential evolution processes, now the device of accidental validation selects two random constraints vectors at random from random constraints pond, it is in random constraints pond, selects the first random constraints vector and the second random constraints vector, calculate the difference vector between the first random constraints vector and the second random vector, then according to weighter factor, to the difference vector weighting between the first random constraints vector and the second random vector, it is and the difference vector between the first random constraints vector and the second random vector is multiplied by weighter factor, sue for peace with the 3rd random constraints vector selected at random in random constraints pond then according to preset variation rule and produce variation vector, it is and determines variation vector.
It should be noted that weighter factor represents the magnification level to difference vector, value should between 0 to 2. Wherein, by experiment and analyze, when the value of weighter factor is more than 1, difference algorithm will be unable to convergence, and value is too little also easily causes too early convergence, and the result obtained is good not. It is therefore preferable that span should be between 0.5 to 1.
Further, the initial value of weighter factor can be set to 0.7.
Need to say, preset variation rule and pre-set, for the rule of each random constraints parameter size in definitive variation vector. User can be configured according to the actual requirements.
By vectorial for variation and the 4th random constraints vector, according to CR (CrossoverRate, crossover probability), carry out mixed processing, obtain trial vector.
Wherein, the 4th random constraints vector is other any one random constraints vector except the first random constraints vector, outside the second random constraints vector and the 3rd random constraints vector in random constraints pond.
Concrete, said process is intersection operation, and now variation vector is carried out mixed processing with the 4th random constraints vector selected at random from random constraints pond by the device of accidental validation, generates trial vector, is acquisition trial vector. Wherein, variation vector and the 4th random constraints vector are carried out mixed processing according to crossover probability CR, be each value in the new experiment vector of generation, be all determined by one of them value of correspondence position in variation vector and the 4th random constraints two vectors of vector. And determine to select the respective value of which vector, a crossover probability CR control, it is possible to by setting CR, it is possible to generate mixed number according to mixed function, according to the relation between mixed number and CR, it is determined which limit is the result of vector mixing be more biased towards. Such as, mixed number, more than CR value, chooses the random constraints parameter in variation vector, and mixing is not more than CR value, selects the random constraints parameter in the 4th random constraints vector.
It should be noted that in the embodiment of the present invention, when generating trial vector, when determining each random constraints parameter of experiment vector, mixed function all needs to generate a random data. That is, when trial vector includes 10 random constraints parameters, mixed function need to generate 10 random data, it is used for 10 random constraints parameters from 10 random constraints parameters and the 4th random constraints vector of variation vector, according to CR value, determine 10 random constraints parameters of correspondence position in trial vector respectively.
Such as, if variation vector comprises 10 random constraints parameters, the 4th random constraints vector comprises 10 random constraints parameters. So, when generating trial vector, when generating first random constraints parameter of trial vector, mixed function generates a random number, it is generation mixed number, and this mixed number and CR value are compared, then when mixed number is more than CR value, first random constraints parameter in variation vector is defined as first random constraints parameter of trial vector. When generating second random constraints parameter of trial vector, mixed function need to regenerate a random number, it is and regenerates mixed number, and mixed number and CR value compare again by this, then when mixed number is not more than CR value, second random constraints parameter in the 4th random constraints vector is defined as second random constraints parameter of trial vector. In like manner, it is possible to from variation vector and the 4th random constraints vector, determine other each random constraints parameters of trial vector.
It should be noted that, determined each random constraints parameter of trial vector by the relation between mixed number and CR for how, the present invention does not limit, can according to during by mixed number more than CR, random constraints parameter in variation vector is defined as the random constraints parameter of trial vector, it is also possible to when mixed number is less than CR, the random constraints parameter in variation vector is defined as the random constraints parameter of trial vector, can also by other means, the invention is not limited in this regard.
Further, the span of CR is 0��1.
Preferably, in order to avoid Premature Convergence. The span of CR is between 0.8 to 1.
Further, the initial value of CR could be arranged to 0.8.
Simulation process is carried out, the coverage rate after acquisition process according to trial vector; Determine that whether the coverage rate after process is more than described initial coverage rate; If the coverage rate after processing is more than initial coverage rate, then according to trial vector, update random constraints pond.
Concrete, above-mentioned steps is for selecting operation. Now, the device of accidental validation is after generating trial vector, it is possible to experiment vector replaces the 4th random constraints vector in random constraints pond, and each random constraints vector in random constraints pond is carried out the simulation process of T time, the coverage rate after being processed. Compare with initial coverage rate to by the coverage rate after process, if the more initial coverage rate of coverage rate after processing is improved, then the 4th random constraints vector before can replacing it by trial vector, update random constraints pond. But, if the more initial coverage rate of coverage rate after processing does not improve, then abandon trial vector.
It should be noted that terminating to require is that user pre-sets according to the actual requirements, it is possible to be the restriction to simulation time, it is also possible to be the restriction to the number of times that differential evolution processes, it is also possible to be the restriction to coverage rate, the invention is not limited in this regard.
Further, terminate to require to include: coverage rate reaches goal-selling coverage rate. It is, terminates the coverage rate after requirement processes and reach goal-selling coverage rate, the differential evolution to random constraints pond could be terminated and process.
Now, random constraints pond is carried out differential evolution process by circulation, and obtains simulated conditions parameter, it is determined that whether simulated conditions parameter meets is terminated, and requires to include until simulated conditions parameter meets end:
Random constraints pond is carried out differential evolution process by circulation, and obtains the coverage rate after process, it is determined that whether the coverage rate after process reaches goal-selling coverage rate, until the coverage rate after processing reaches goal-selling coverage rate.
That is, the device of accidental validation is after carrying out differential evolution process to random constraints pond, obtain random constraints pond and carry out the coverage rate after differential evolution processes, and the coverage rate after processing is compared with goal-selling coverage rate, it is determined that whether the coverage rate after process reaches goal-selling coverage rate. When coverage rate after treatment reaches goal-selling coverage rate, illustrating that loop stop conditions meets, now, the device of accidental validation is without carrying out differential evolution process to random constraints pond. When coverage rate after treatment is not reaching to goal-selling coverage rate, now, the device of accidental validation needs to continue random constraints pond is carried out differential evolution process, and reacquire the coverage rate after differential evolution processes, coverage rate after the process of this reacquisition and goal-selling coverage rate are contrasted, when determining that coverage rate is not reaching to goal-selling coverage rate, need again again random constraints pond to be carried out differential evolution process, until the coverage rate after processing meets goal-selling coverage rate, it is loop stop conditions and meets position.
Further, when loop stop conditions is unsatisfactory for, the weighter factor of mutation operation in differential evolution can also being processed, and the CR intersected in operation is adjusted, so that random constraints pond can be carried out differential evolution process according to the weighter factor after adjusting and CR by the device of accidental validation.
Now, as it is shown on figure 3, random constraints pond is carried out differential evolution process by circulation, and simulated conditions parameter is obtained, it is determined that whether simulated conditions parameter meets is terminated, and requires to include until simulated conditions parameter meets end:
Random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, it is determined that whether simulated conditions parameter meets and terminates requirement, and simulated conditions parameter be unsatisfactory for end require time, update weighter factor and CR, until simulated conditions parameter meets terminates requirement.
It is to say, the device of accidental validation is after carrying out differential evolution process to random constraints pond, it is possible to obtain simulated conditions parameter, and then simulated conditions parameter is contrasted with terminating requirement, it is determined that whether simulated conditions parameter meets is terminated requirement. and simulated conditions parameter meet end require time, no longer random constraints pond is carried out differential evolution process, it is possible to perform next step. and simulated conditions parameter be unsatisfactory for end require time, weighter factor and CR can be adjusted, and according to the weighter factor after adjusting and CR, random constraints pond is carried out differential evolution process, obtain and reacquire simulated conditions parameter, determine whether the simulated conditions parameter of reacquisition meets and terminate requirement, and reacquire simulated conditions parameter be unsatisfactory for end require time, readjust weighter factor and CR, the device of accidental validation is according to the weighter factor readjusted and CR, random constraints pond is carried out differential evolution process, until simulated conditions parameter meets terminates requirement.
It should be noted that renewal weighter factor and CR can be the methods that user sets renewal according to the actual requirements, it is also possible to be that the coverage rate after the process according to acquisition every time adjusts weighter factor and CR, the invention is not limited in this regard.
It should be noted that, the device of accidental validation is when determining that initial coverage rate reaches goal-selling coverage rate, illustrate that each random constraints vector in random constraints pond is more excellent, now, can no longer perform step 103, but the random constraints vector directly utilized in random constraints pond carries out accidental validation.
Step 104, according to differential evolution process after random constraints pond in random constraints vector, carry out accidental validation.
Concrete, the device of accidental validation, after random constraints pond has carried out differential evolution process, obtains the random constraints pond after differential evolution processes. Now, the device of accidental validation can utilize each random constraints vector in the random constraints pond after differential evolution process, carries out corresponding accidental validation.
It should be noted that how the device of accidental validation carries out accidental validation according to random constraints vector is prior art, the invention is not limited in this regard.
A kind of method embodiments providing accidental validation, comprises determining that out at least four random constraints vector comprised in random constraints pond; Respectively the random constraints vector in random constraints pond is carried out simulation process, obtain initial coverage rate; When initial coverage rate does not arrive goal-selling coverage rate, random constraints pond is carried out differential evolution process by circulation, and obtains simulated conditions parameter, it is determined that whether simulated conditions parameter meets is terminated, until simulated conditions parameter meets terminates requirement; Terminate to require to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process; Random constraints vector in described random constraints pond after processing according to differential evolution, carries out accidental validation. So, in the accidental validation that random constraints vector is more complicated, it is achieved that the self-adaptative adjustment to random constraints vector. Under the premise that need not increase labor workload, can emulate with different good random constraints vectors all the time, solve the waste of time that long unsuitable random constraints vector brings and resource, thus drastically increasing the efficiency of accidental validation. And then achieve while ensureing verification the verifying results, improve the purpose of verification efficiency.
Embodiments provide the device of a kind of accidental validation, as shown in Figure 4, including:
Determine unit 301, for determining at least four random constraints vector comprised in random constraints pond.
Wherein, random constraints vector includes at least one random constraints parameter.
Processing unit 302, for respectively the random constraints vector in random constraints pond being carried out simulation process, obtains initial coverage rate.
Processing unit 302, being additionally operable to when initial coverage rate is not up to goal-selling coverage rate, random constraints pond is carried out differential evolution process by circulation, and obtains simulated conditions parameter, determine whether simulated conditions parameter meets and terminate requirement, until simulated conditions parameter meets terminates requirement.
Wherein. Terminate to require to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process.
Concrete, processing unit 302, specifically for selecting the first random constraints vector and the second random constraints vector in random constraints pond, calculate the difference vector of described first random constraints vector and the second random constraints vector chosen; It is weighted processing to difference vector according to weighter factor, and according to default variation rule, utilizes the difference vector after weighting and the 3rd random constraints vector, it is determined that variation vector.
By vectorial for variation and the 4th random constraints vector, according to crossover probability CR, carry out mixed processing, obtain trial vector.
Simulation process is carried out, the coverage rate after acquisition process according to trial vector.
Determine that whether the coverage rate after process is more than initial coverage rate.
If the coverage rate after processing is more than initial coverage rate, then according to trial vector, update random constraints pond.
Wherein, the first random constraints vector and the second random constraints vector are any two random constraints vectors in random constraints pond. 3rd random constraints vector is other any one the random constraints vectors except the first random constraints vector and the second random constraints vector in random constraints pond. 4th random constraints vector is other any one random constraints vector except the first random constraints vector, outside the second random constraints vector and the 3rd random constraints vector in random constraints pond.
Further, terminate to require to include: coverage rate reaches described goal-selling coverage rate.
Now, processing unit 302, specifically for circulating, random constraints pond is carried out differential evolution process, and obtain the coverage rate after process, it is determined that whether the coverage rate after process reaches goal-selling coverage rate, until the coverage rate after processing reaches goal-selling coverage rate.
Further, processing unit 302, specifically for circulation, random constraints pond is carried out differential evolution process, and obtain simulated conditions parameter, determine whether simulated conditions parameter meets and terminate requirement, and simulated conditions parameter be unsatisfactory for end require time, update weighter factor and CR, until simulated conditions parameter meets terminates requirement.
Further, the initial value of weighter factor is 0.7. The initial value of CR is 0.8.
Authentication unit 303, for the random constraints vector in the random constraints pond after processing according to differential evolution, carries out accidental validation.
Embodiments provide the device of a kind of accidental validation, comprise determining that out at least four random constraints vector comprised in random constraints pond; Respectively the random constraints vector in random constraints pond is carried out simulation process, obtain initial coverage rate; When initial coverage rate does not arrive goal-selling coverage rate, random constraints pond is carried out differential evolution process by circulation, and obtains simulated conditions parameter, it is determined that whether simulated conditions parameter meets is terminated, until simulated conditions parameter meets terminates requirement; Terminate to require to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process; Random constraints vector in described random constraints pond after processing according to differential evolution, carries out accidental validation. So, in the accidental validation that random constraints vector is more complicated, it is achieved that the self-adaptative adjustment to random constraints vector. Under the premise that need not increase labor workload, can emulate with different good random constraints vectors all the time, solve the waste of time that long unsuitable random constraints vector brings and resource, thus drastically increasing the efficiency of accidental validation. And then achieve while ensureing verification the verifying results, improve the purpose of verification efficiency.
Last it is noted that above example is only in order to illustrate technical scheme, it is not intended to limit;Although the present invention being described in detail with reference to previous embodiment, it will be understood by those within the art that: the technical scheme described in foregoing embodiments still can be modified by it, or wherein portion of techniques feature is carried out equivalent replacement; And these amendments or replacement, do not make the essence of appropriate technical solution depart from the spirit and scope of various embodiments of the present invention technical scheme.
Claims (10)
1. the method for an accidental validation, it is characterised in that including:
Determine at least four random constraints vector comprised in random constraints pond; Described random constraints vector includes at least one random constraints parameter;
Respectively the random constraints vector in described random constraints pond is carried out simulation process, obtain initial coverage rate;
When described initial coverage rate is not up to goal-selling coverage rate, described random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, it is determined that whether described simulated conditions parameter meets is terminated requirement, until simulated conditions parameter meets terminates requirement; Described end requires to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process;
Random constraints vector in described random constraints pond after processing according to differential evolution, carries out accidental validation.
2. method according to claim 1, it is characterised in that
Described described random constraints pond is carried out differential evolution process include:
Described random constraints pond selects the first random constraints vector and the second random constraints vector, calculates the difference vector of described first random constraints vector and the second random constraints vector chosen; It is weighted processing to described difference vector according to weighter factor, and according to default variation rule, utilizes the described difference vector after weighting and the 3rd random constraints vector, it is determined that variation vector; Described first random constraints vector and the second random constraints vector are any two random constraints vectors in described random constraints pond; Described 3rd random constraints vector is other any one the random constraints vectors except described first random constraints vector and the second random constraints vector in described random constraints pond;
By described variation vector and the 4th random constraints vector, according to crossover probability CR, carry out mixed processing, obtain trial vector; Described 4th random constraints vector is other any one random constraints vector except described first random constraints vector, outside the second random constraints vector and the 3rd random constraints vector in described random constraints pond;
Simulation process is carried out, the coverage rate after acquisition process according to described trial vector;
Determine that whether the coverage rate after described process is more than described initial coverage rate;
If the coverage rate after described process is more than described initial coverage rate, then according to described trial vector, update described random constraints pond.
3. method according to claim 2, it is characterised in that
Described end requires to include: coverage rate reaches described goal-selling coverage rate;
Described random constraints pond is carried out differential evolution process by described circulation, and obtains simulated conditions parameter, it is determined that whether described simulated conditions parameter meets is terminated, and requires to include until simulated conditions parameter meets end:
Described random constraints pond is carried out differential evolution process by circulation, and obtains the coverage rate after process, it is determined that whether the coverage rate after described process reaches goal-selling coverage rate, until the coverage rate after processing reaches described goal-selling coverage rate.
4. according to the method in claim 2 or 3, it is characterized in that, described random constraints pond is carried out differential evolution process by described circulation, and obtains simulated conditions parameter, determine whether described simulated conditions parameter meets to terminate, require to include until simulated conditions parameter meets end:
Described random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, it is determined that whether described simulated conditions parameter meets and terminates requirement, and simulated conditions parameter be unsatisfactory for end require time, update weighter factor and CR, until simulated conditions parameter meets terminates requirement.
5. method according to claim 4, it is characterised in that
The initial value of described weighter factor is 0.7; The initial value of described CR is 0.8.
6. the device of an accidental validation, it is characterised in that including:
Determine unit, for determining at least four random constraints vector comprised in random constraints pond; Described random constraints vector includes at least one random constraints parameter;
Processing unit, for respectively the random constraints vector in described random constraints pond being carried out simulation process, obtains initial coverage rate;
Described processing unit, it is additionally operable to when described initial coverage rate is not up to goal-selling coverage rate, described random constraints pond is carried out differential evolution process by circulation, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets and terminate requirement, until simulated conditions parameter meets terminates requirement; Described end requires to pre-set, for terminating random constraints pond is carried out the requirement information of differential evolution process;
Authentication unit, for the random constraints vector in the described random constraints pond after processing according to differential evolution, carries out accidental validation.
7. device according to claim 6, it is characterised in that
Described processing unit, specifically for selecting the first random constraints vector and the second random constraints vector in described random constraints pond, calculates the difference vector of described first random constraints vector and the second random constraints vector chosen; It is weighted processing to described difference vector according to weighter factor, and according to default variation rule, utilizes the described difference vector after weighting and the 3rd random constraints vector, it is determined that variation vector; Described first random constraints vector and the second random constraints vector are any two random constraints vectors in described random constraints pond; Described 3rd random constraints vector is other any one the random constraints vectors except described first random constraints vector and the second random constraints vector in described random constraints pond;
By described variation vector and the 4th random constraints vector, according to crossover probability CR, carry out mixed processing, obtain trial vector; Described 4th random constraints vector is other any one random constraints vector except described first random constraints vector, outside the second random constraints vector and the 3rd random constraints vector in described random constraints pond;
Simulation process is carried out, the coverage rate after acquisition process according to described trial vector;
Determine that whether the coverage rate after described process is more than described initial coverage rate;
If the coverage rate after described process is more than described initial coverage rate, then according to described trial vector, update described random constraints pond.
8. device according to claim 7, it is characterised in that
Described end requires to include: coverage rate reaches described goal-selling coverage rate;
Described processing unit, specifically for circulation, described random constraints pond is carried out differential evolution process, and obtain the coverage rate after process, it is determined that whether the coverage rate after described process reaches goal-selling coverage rate, until the coverage rate after processing reaches described goal-selling coverage rate.
9. the device according to claim 7 or 8, it is characterised in that
Described processing unit, specifically for circulation, described random constraints pond is carried out differential evolution process, and obtain simulated conditions parameter, determine whether described simulated conditions parameter meets and terminate requirement, and simulated conditions parameter be unsatisfactory for end require time, update weighter factor and CR, until simulated conditions parameter meets terminates requirement.
10. the device according to claim 7 or 8, it is characterised in that
The initial value of described weighter factor is 0.7; The initial value of described CR is 0.8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511019072.1A CN105653793B (en) | 2015-12-29 | 2015-12-29 | A kind of method and device of accidental validation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511019072.1A CN105653793B (en) | 2015-12-29 | 2015-12-29 | A kind of method and device of accidental validation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105653793A true CN105653793A (en) | 2016-06-08 |
| CN105653793B CN105653793B (en) | 2019-03-05 |
Family
ID=56477445
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511019072.1A Active CN105653793B (en) | 2015-12-29 | 2015-12-29 | A kind of method and device of accidental validation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105653793B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106202638A (en) * | 2016-06-29 | 2016-12-07 | 醴陵恒茂电子科技有限公司 | Chip verification platform emulation ending control method and system |
| CN107193738A (en) * | 2017-05-19 | 2017-09-22 | 郑州云海信息技术有限公司 | A kind of verification method that dynamic adjustment random constraints are covered based on function point |
| CN107273604A (en) * | 2017-06-12 | 2017-10-20 | 郑州云海信息技术有限公司 | The selectable random verification method based on Coverage- Driven of random constraints |
| CN107908514A (en) * | 2017-11-15 | 2018-04-13 | 北京中电华大电子设计有限责任公司 | A kind of emulation verification method of power consumption adaptive circuit |
| CN113051149A (en) * | 2019-12-27 | 2021-06-29 | 深圳云天励飞技术有限公司 | Case parameter generation method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150149131A1 (en) * | 2013-11-25 | 2015-05-28 | Sony Corporation | Simulation method, simulation program, processing apparatus, simulator, and design method |
| CN104679634A (en) * | 2015-03-06 | 2015-06-03 | 浪潮电子信息产业股份有限公司 | Self-adapted random verification method used for simulating ultra-large-scale chip |
| CN105005646A (en) * | 2015-07-02 | 2015-10-28 | 中国科学院力学研究所 | Holographic method for analyzing stability safety of side slope |
-
2015
- 2015-12-29 CN CN201511019072.1A patent/CN105653793B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150149131A1 (en) * | 2013-11-25 | 2015-05-28 | Sony Corporation | Simulation method, simulation program, processing apparatus, simulator, and design method |
| CN104679634A (en) * | 2015-03-06 | 2015-06-03 | 浪潮电子信息产业股份有限公司 | Self-adapted random verification method used for simulating ultra-large-scale chip |
| CN105005646A (en) * | 2015-07-02 | 2015-10-28 | 中国科学院力学研究所 | Holographic method for analyzing stability safety of side slope |
Non-Patent Citations (2)
| Title |
|---|
| 李拓 等: "一种Cache一致性协议验证中覆盖率驱动的随机验证方法", 《计算机应用与软件》 * |
| 詹瑾瑜 等: "系统级芯片测试调度最优总线指定方法", 《计算机集成制造系统》 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106202638A (en) * | 2016-06-29 | 2016-12-07 | 醴陵恒茂电子科技有限公司 | Chip verification platform emulation ending control method and system |
| CN106202638B (en) * | 2016-06-29 | 2019-04-23 | 湖南恒茂高科股份有限公司 | Chip verification platform emulates ending control method and system |
| CN107193738A (en) * | 2017-05-19 | 2017-09-22 | 郑州云海信息技术有限公司 | A kind of verification method that dynamic adjustment random constraints are covered based on function point |
| CN107273604A (en) * | 2017-06-12 | 2017-10-20 | 郑州云海信息技术有限公司 | The selectable random verification method based on Coverage- Driven of random constraints |
| CN107273604B (en) * | 2017-06-12 | 2020-09-22 | 苏州浪潮智能科技有限公司 | Random constraint selectable coverage rate driving-based random verification method |
| CN107908514A (en) * | 2017-11-15 | 2018-04-13 | 北京中电华大电子设计有限责任公司 | A kind of emulation verification method of power consumption adaptive circuit |
| CN113051149A (en) * | 2019-12-27 | 2021-06-29 | 深圳云天励飞技术有限公司 | Case parameter generation method and device |
| CN113051149B (en) * | 2019-12-27 | 2023-11-17 | 深圳云天励飞技术有限公司 | Use case parameter generation method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105653793B (en) | 2019-03-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105653793A (en) | Random verification method and apparatus | |
| Rudolph et al. | Capabilities of EMOA to detect and preserve equivalent Pareto subsets | |
| US20220182406A1 (en) | Analysis apparatus, analysis system, analysis method, and non-transitory computer readable medium storing program | |
| CN108900333A (en) | A kind of appraisal procedure and assessment device of quality of wireless network | |
| Rangel et al. | On redundant coverage maximization in wireless visual sensor networks: Evolutionary algorithms for multi-objective optimization | |
| CN105739956B (en) | The method and system of the building intelligent rules model of computer system | |
| CN110135582A (en) | Neural metwork training, image processing method and device, storage medium | |
| CN105808426A (en) | Path coverage test data generation method used for weak mutation test | |
| CN104679634A (en) | Self-adapted random verification method used for simulating ultra-large-scale chip | |
| CN112181698A (en) | Method, device and equipment for testing automatic driving limit performance and storage medium | |
| CN108563875A (en) | Analog circuit measuring point and frequency based on multiple-objection optimization combine preferred method | |
| CN110046638A (en) | Fusion method, device and the equipment of multi-platform data | |
| CN101853202B (en) | Test case autogeneration method based on genetic algorithm and weighted matching algorithm | |
| CN106874813B (en) | A kind of method automatically selecting M1 card and smart card | |
| CN112988606B (en) | Test case generation method, device, equipment and storage medium | |
| CN110222816A (en) | Method for building up, image processing method and the device of deep learning model | |
| CN110210623A (en) | Adaptive multiple target hybrid differential evolution algorithm based on simulated annealing and comentropy | |
| Hale et al. | Optimality versus run time for isolated signalized intersections | |
| CN106156107A (en) | A kind of discovery method of hot news | |
| CN109522679B (en) | Method and system for generating functional excitation vector based on classification processing | |
| JP2020021301A (en) | Training data evaluation device, training data evaluation method, and program | |
| US20220052897A1 (en) | Selection system for waveforms and waveform parameters in 5g and beyond next generation communication systems | |
| Józefczyk et al. | Scatter Search based algorithms for min-max regret task scheduling problems with interval uncertainty | |
| CN109978146A (en) | A kind of super ginseng optimization method and system of convolutional neural networks | |
| Theodossiou et al. | Harmony Search Algorithm, a novel optimization technique. Parameter calibration and applications on water resources management problems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |