CN107478981A - A kind of test and excitation set method for optimizing based on greedy algorithm - Google Patents
A kind of test and excitation set method for optimizing based on greedy algorithm Download PDFInfo
- Publication number
- CN107478981A CN107478981A CN201710725698.7A CN201710725698A CN107478981A CN 107478981 A CN107478981 A CN 107478981A CN 201710725698 A CN201710725698 A CN 201710725698A CN 107478981 A CN107478981 A CN 107478981A
- Authority
- CN
- China
- Prior art keywords
- test
- test excitation
- new
- excitations
- excitation
- 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
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/316—Testing of analog circuits
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Abstract
A kind of test and excitation set method for optimizing based on greedy algorithm, the present invention relates to test and excitation set method for optimizing.The present invention in order to solve influence of parameter error degree of the prior art in component to circuit it is smaller when, fault detect rate and diagnosis are relatively low, and the problem of can not cover the component of whole circuit.The present invention includes：Step 1：Establish the initial matrix that m test and excitations to be selected are formed with the component in n circuit；Step 2：The initial matrix that is obtained according to step 1 simultaneously utilizes greedy algorithm, preferably goes out x test and excitation, T from m test and excitations to be selected successively1, T2, T3…Tx, constitute test and excitation set F, F=(T1, T2, T3…Tx), obtain minimum cost corresponding to test and excitation set F and be Smin；Step 3：Optimized using internal comparison strategy and the random test test and excitation set F for rejecting strategy to step 2.The present invention is used for analog circuit fault diagnosing field.
Description
Technical Field
The invention relates to a greedy algorithm based test stimulus set optimization method.
Background
With the development of national defense science and technology, electronic systems are widely applied to the fields of missile control, communication, target detection, friend or foe identification and the like, so that the reliability of the electronic systems determines the performance of weaponry. Although the proportion of the analog circuit of the electronic system is about 20%, most faults of the electronic system are from the analog circuit, one of main factors influencing the performance of the analog circuit is faults caused by parameter drift of components in the circuit, most researches are carried out in the field of current fault diagnosis aiming at the faults, in order to improve detection and diagnosis of the faults, test excitation is carried out on the faults in the circuit, and all components in the analog circuit can find parameter deviations of different degrees under the dual action of an external environment and an internal environment to cause the faults of the analog circuit, so a test excitation set capable of covering the whole circuit is required to be provided, and the faults caused by different deviations of the components of the circuit can be effectively excited to the maximum extent.
For a certain fault in the circuit, under excitation of different frequencies, the fault output response is different, so that a proper test excitation is selected to obtain the optimal circuit output response of the fault, and the obtained fault sample is separable as much as possible. Reasonable test excitation can optimize the fault resolution and is easy for the classifier to detect faults.
Disclosure of Invention
The invention provides a test excitation set optimization method based on a greedy algorithm, which aims to solve the problems that a circuit test excitation set obtained in the prior art cannot effectively excite faults caused by different parameter deviation ratios of components in a circuit, the fault detection rate and the diagnosis rate are lower when the parameter deviation degree of the components has small influence on the circuit, and the existing test excitation set cannot cover the components of the whole circuit.
A greedy algorithm based test stimulus set optimization method comprises the following steps:
the method comprises the following steps: establishing an initial matrix formed by m to-be-selected test excitations and components in n circuits, wherein the cost of each to-be-selected test excitation is set as s 1 ,s 2 ,...,s i ,...,s m ；
Step two: the optimal optimization target of the initial matrix and test stimulus set obtained according to the step one is to enable the set to cover all n devices, meanwhile, the sum of the corresponding costs of the set is minimum, so the problem is solved by adopting a set covering problem with a weight value, x test stimuli are sequentially and preferably selected from m test stimuli to be selected by utilizing a greedy algorithm, and T test stimuli are selected 1 ，T 2 ，T 3 …T x Forming a test excitation set F, F = (T) 1 ，T 2 ，T 3 …T x ) Obtaining the minimum cost sum S corresponding to the test excitation set F min ；
Step three: and (4) optimizing the test excitation set F in the step two by adopting an internal comparison strategy and a random elimination strategy, and improving the goodness of the test excitation set.
The invention provides a test excitation set optimization method based on a greedy algorithm, which can optimize a test excitation set which covers the whole circuit element and has a better excitation effect.
Firstly, an initial matrix of the condition that test excitations cover circuit components is established, which is called an initial pattern in the problem of aggregate coverage, wherein m test excitations to be selected are totally and respectively f 1 ，f 2 ，f 3 …f m The total number of test excitations in the circuit is n, which are respectively H 1 ，H 2 ，H 3 …H n Then the test excitation and the device to be selected form an initial matrix of m × n, and the matrix is shown in table 1, where each element in the matrix is a ij This indicates that i = (1,2,3 \8230; m), and j = (1,2,3 \8230; n). Giving the coverage condition of the test excitation to be selected on the component in the matrix, and when the ith test excitation can excite the jth component, then adding the element a at the corresponding position in the initial matrix ij When the ith test excitation can not excite the jth element, the element a of the corresponding position in the initial matrix is marked as' 1 ij Noted as "0", the initial matrix is a Boolean matrix of 0-1, where each test stimulus has a corresponding cost value of s 1 ，s 2 ，s 3 …s n As shown in table 1;
TABLE 1
The idea of the greedy algorithm is to select rows from m rows in turn according to some rule to cover all columns, so that the sum of the costs of the selected rows is as small as possible. At a certain moment, rows have been selected according to the selection rule to cover columns, and columns are not covered, which is called a pattern. For convenience of presentation, we use an n-dimensional boolean vector P to represent the coverage of the column. Under a certain pattern, if the jth column is already covered, P j = true, otherwise P j = false (j =1,2,3, \8230n). For any row i, its average cost is c i On this basis, the average cost of i under the current configuration can be defined by the following formula (1):
the greedy algorithm selection strategy may be described as selecting the average cost minimum c from the initial constellation i To cover those columns that have not been covered, and select according to this strategyAfter one row, a new pattern is obtained, wherein the test stimulus corresponding to the row with the minimum average cost is selected as an element T in the final preferred test stimulus set 1 (ii) a Starting from the new pattern, the greedy algorithm selection strategy is used again until all columns are covered or no rows in the remaining rows can cover the columns which are not necessarily covered by the row, and the elements T in the optimal test stimulus set F are selected in turn based on the selection strategy 1 ，T 2 ，T 3 …T x The sum of the costs corresponding to the set is S min 。
Aiming at the obtained test excitation set F, the method adopts two optimization strategies, namely an internal comparison strategy and a random elimination strategy to further optimize the test excitation set F. The optimization strategy is to compare in the test excitation set and to a certain test T in the set x Covering p components, if T in the available set x The test stimulus beyond that covers p components, T can be converted to x Removing from the set; if there is no T in the usable set x If the test stimulus covers p components, T cannot be connected x And (5) removing from the set.
The second optimization strategy is a random elimination strategy, and the specific process is as follows: the number of times of rejecting test stimulus to the set F is set as kFirstly, randomly removing any two test excitations from a set F, which can result in that q components are not covered; selecting test excitation from the remaining test excitation to be selected according to a greedy algorithm strategy to cover q components to form a new test excitation set F new Calculate the set F new Cost and S of new With the cost and S of the set F min Comparing; if S is min <S new If the set F is the optimal test excitation set; if S is min >S new Then use the new set F new Replacing F to become an optimal test excitation set; removing and optimizing the excitation in the set F according to the strategy until the removal times k reachSub or F new And (5) obtaining an optimal test excitation set in the same way as the F, and optimizing the test excitation set F to the maximum extent.
The invention has the beneficial effects that:
the invention aims to solve the problems that in the existing analog circuit, due to the fact that different degrees of parameter deviation occur to devices under the action of working environment, circuit faults are caused, the influence degree of the parameter deviation degree of the devices on the circuit is different, the influence is large, effective detection can be achieved, and if the influence is small, the detection is not easy. Therefore, in order to improve the fault detection rate and the diagnosis rate caused by the deviation of the components of the whole circuit, the invention provides a test excitation set optimization method based on a greedy algorithm, and the test excitation set optimized by the method can cover the components of the whole circuit.
Therefore, the invention mainly aims at the analog circuit fault caused by the deviation of the device to carry out test excitation optimization, and effectively improves the detection rate and the diagnosis rate of the analog circuit fault. The test excitation set obtained by the greedy algorithm-based test excitation set optimization method can cover faults caused by component parameter deviation of the whole analog circuit, meanwhile, better excitation effect on the faults is guaranteed, the faults caused by different parameter deviation of the components have higher resolution, fault detection is carried out by combining a classifier, the detection rate reaches over 95%, the faults caused by smaller parameter deviation also have better excitation effect, and the detection rate reaches over 92%, so that the effect which is not achieved by the traditional test excitation selection method at present is achieved.
Drawings
Fig. 1 is a circuit diagram of a Leapfrog circuit.
Detailed Description
The first embodiment is as follows: a greedy algorithm-based test excitation set optimization method comprises the following steps:
the method comprises the following steps: establishing an initial matrix formed by m to-be-selected test excitations and components in n circuits, wherein the cost of each to-be-selected test excitation is set as s 1 ,s 2 ,...,s i ,...,s m ；
Step two: according to the initial matrix obtained in the step one and by utilizing a greedy algorithm, the optimal selection target of the test excitation set is to enable the set to cover all n devices, meanwhile, the sum of the corresponding costs of the set is minimum, so that the problem is solved by adopting a set covering problem with a weight value, x test excitations are selected from m to-be-selected test excitations in sequence, and T test excitations are selected 1 ，T 2 ，T 3 …T x A set of test excitations F, F = (T) is formed 1 ，T 2 ，T 3 …T x ) Obtaining the minimum cost sum S corresponding to the test excitation set F min ；
Step three: and (4) optimizing the test excitation set F in the step two by adopting an internal comparison strategy and a random elimination strategy, and improving the goodness of the test excitation set.
Aiming at faults in the circuit, the invention provides a test excitation set optimization method based on a greedy algorithm, which is used for converting a test excitation optimization problem into a set coverage problem with weights and searching a test excitation set capable of covering the whole circuit and ensuring a better excitation effect on the faults.
The second embodiment is as follows: the first difference between the present embodiment and the specific embodiment is: the specific process of establishing an initial matrix formed by the m test excitations to be selected and the components in the circuit in the first step is as follows:
the method comprises the following steps: m candidate test excitations are respectively f 1 ，f 2 ，f 3 …f m The components in the n circuits are respectively H 1 ，H 2 ，H 3 …H n To be selectedThe test excitation and the components form an m multiplied by n empty matrix;
the first step is: step one, each element in the empty matrix is a ij I =1,2,3 \ 8230m, j =1,2,3 \ 8230n; giving the coverage condition of the test excitation to be selected on the components in the matrix, and when the test excitation in the ith row can excite the components in the jth column, giving the element a at the corresponding position in the empty matrix ij When the ith row test excitation can not excite the jth column element, the element a of the corresponding position is marked as 1 ij And marking as 0, and obtaining a matrix formed by the to-be-selected test excitation and the element, wherein the initial matrix is 0-1.
Other steps and parameters are the same as those in the first embodiment.
The third concrete implementation mode: the present embodiment differs from the first or second embodiment in that: in the second step, x test stimuli T are sequentially selected from m to-be-selected test stimuli according to the initial matrix obtained in the first step by using a greedy algorithm 1 ，T 2 ，T 3 …T x A set of test excitations F, F = (T) is formed 1 ，T 2 ，T 3 …T x ) Obtaining the minimum cost sum S corresponding to the test excitation set F min The specific process comprises the following steps:
step two is as follows: according to the initial matrix obtained in the step one, the excitation effect of the test excitation on the component is represented by an n-dimensional Boolean vector P, and if the jth column is excited, the P is j True, otherwise P j = false, and the average cost c of each of m lines is calculated using equation (1) separately 1 ,c 2 ,...,c i ,...,c m
Step two: according to the calculation result of the average cost of each line in the second step, a selection strategy is made by a greedy algorithm, and an average generation is selectedExciting the not excited components by the test excitation with the minimum price calculation result, and then taking the test excitation with the minimum average price calculation result as the first element T in the selected test excitation set F 1 ；
Step two and step three: removing the excited components from the initial matrix to form a new matrix; re-executing the first step and the second step to select the second element T in the test excitation set F from the new matrix 2 Until all components in the circuit are excited, x test excitations are preferably selected to form a test excitation set F, F = (T) exciting the whole circuit 1 ，T 2 ，T 3 …T x ) And the minimum cost sum corresponding to the test excitation set F is S min ＝s 1 +s 2 +…+s x 。
Other steps and parameters are the same as those in the first or second embodiment.
The fourth concrete implementation mode is as follows: the difference between this embodiment mode and one of the first to third embodiment modes is: the specific process of optimizing the test excitation set F in the second step by adopting an internal comparison strategy and a random elimination strategy in the third step is as follows:
step three, first: and optimizing the test excitation set F by adopting an internal comparison strategy, namely: test stimulus T in test stimulus set F w P components are excited, w is 1,2, \ 8230;, x, if T is divided by T in the test excitation set F w If other test stimuli can excite the same p components, T will be w Removing from the test excitation set F; if T is divided in the test excitation set F w If the other test stimuli cannot excite the same p components, T cannot be converted to w Removing from the set; obtaining an optimized test excitation set F;
step three: optimizing the optimized test excitation set F obtained in the third step by adopting a random elimination strategy; namely: randomly removing any two test excitations from the x test excitations of the test excitation set F each time, wherein the total removing times are k times,forming a new test excitation set after each elimination and recording the new test excitation set as F new The sum of the costs is S new . Removing any two test excitations from the optimized test excitation set F, generating q elements which are not excited, executing the step two of selecting the test excitations for exciting the q elements from the remaining m-x test excitations to be selected, and forming a new test excitation set F by the selected test excitations and the test excitations which are not removed from the test excitation set F new Calculating a new set of test stimuli F new Cost and S of new ；
Step three: new set of test stimuli F new Cost and S of new Cost sum S with set F min Comparison, if S min <S new If the set F is the optimal test excitation set; if S is min >S 1 Then use the new set F new Replacing F to be an optimal test excitation set after the first test excitation elimination; re-executing step three or two (test excitation of each rejection is not identical) until the rejection number k reachesSub or F new And obtaining an optimal test excitation set in the same way as the F.
Other steps and parameters are the same as those in one of the first to third embodiments.
The first embodiment is as follows:
the invention takes a Leapfragg circuit as an example to explain the test excitation set optimization method based on the greedy algorithm in detail. Firstly, a simulation circuit is built in the PSPice software as shown in fig. 1. In the figure, R1= R2= R3= R4= R5= R6= R7= R8= R9= R10= R11= R12= R13=10k Ω, C1= C4=10nf, C2= C3=20nf, where the tolerance of the resistance and the capacitance is 5%. The circuit has 17 components, the optimal test excitation of each component is obtained correspondingly, and the coverage of the obtained test excitation on the components is shown in table 2.
TABLE 2 Leapfrog circuit candidate test excitation and element initial corresponding matrix
| C 1 | C 2 | C 3 | C 4 | R 1 | R 2 | R 3 | R 4 | R 5 | R 6 | R 7 | R 8 | R 9 | R 10 | R 11 | R 12 | R 13 | Cost of | |
| f 1 | 1 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0.03647 |
| f 2 | 1 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.28310 |
| f 3 | 1 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.26462 |
| f 4 | 1 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0.02514 |
| f 5 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0.19156 |
| f 6 | 1 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0.02256 |
| f 7 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 1 | 1 | 0.45460 |
| f 8 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 1 | 1 | 0.45376 |
| f 9 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.05570 |
| f 10 | 1 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0.30085 |
| f 11 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0.27908 |
| f 12 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 0.44262 |
| f 13 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 0 | 0.14313 |
| f 14 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 0.43995 |
| f 15 | 1 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0.02256 |
| f 16 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 0.45215 |
| f 17 | 0 | 0 | 0 | 0 | 1 | 0 | 1 | 1 | 0 | 0 | 0 | 1 | 1 | 1 | 0 | 1 | 1 | 0.44407 |
The greedy algorithm-based test excitation set optimization strategy and the two set optimization strategies provided by the invention obtain a test excitation set with (f) 6 ，f 9 ，f 5 ，f 14 ，f 3 ) The sum of the corresponding costs is 0.9743. In order to verify that the selected test excitation can cover the whole circuit and effectively improve the fault detection rate, the method adopts a support vector data description method (SVDD) as a classifier to carry out fault detection, because the algorithm of the invention mainly aims at the fault caused by each deviation degree of the parameters of the components, the deviation percentages of different parameters of the whole circuit device are used as the fault to be injected into the circuit, and the detection result is shown in Table 3.
TABLE 3 Leapfragg circuit element fault detection rate
Analyzing the above experimental results can obtain: it can be known from table 2 that the failure detection rate caused by different parameter deviation ratios of each device is higher and reaches more than 92.25%, the average detection rate reaches 95.78%, the method also has very good excitation effect on the smaller parameter deviation of the device, such as the failure caused by C2 and R2, and the failure detection rate reaches 93.25% and 93.65%, which fully indicates that the preferred test excitation set of the invention has very good excitation effect on the failure caused by the parameter deviation of different devices in the circuit, and that the higher detection rate is obtained for each device to indicate that the test excitation covers the whole circuit.
The present invention is capable of other embodiments and its several details are capable of modifications in various obvious respects, all without departing from the spirit and scope of the present invention.
Claims (4)
1. A greedy algorithm-based test excitation set optimization method is characterized by comprising the following steps: the greedy algorithm-based test excitation set optimization method comprises the following steps of:
the method comprises the following steps: establishing an initial matrix formed by m to-be-selected test excitations and components in n circuits, wherein the cost of each to-be-selected test excitation is set as s 1 ,s 2 ,...,s i ,...,s m ；
Step two: according to the initial matrix obtained in the first step, x test excitations are selected out of m test excitations to be selected in sequence by means of a greedy algorithm, and T 1 ，T 2 ，T 3 ...T x A set of test excitations F, F = (T) is formed 1 ，T 2 ，T 3 ...T x ) Obtaining the minimum cost sum S corresponding to the test excitation set F min ；
Step three: and optimizing the test excitation set F in the second step by adopting an internal comparison strategy and a random elimination strategy.
2. The greedy algorithm based test stimulus set optimization method of claim 1, wherein: the specific process of establishing an initial matrix formed by the m test excitations to be selected and the components in the circuit in the first step is as follows:
the method comprises the following steps: m candidate test excitations are respectively f 1 ，f 2 ，f 3 ...f m The components in the n circuits are respectively H 1 ，H 2 ，H 3 ...H n The test excitation to be selected and the components form an m multiplied by n empty matrix;
the first step is: step one, each element in the empty matrix is a ij Denotes i =1,2,3.. M, j =1,2,3.. N; when the ith row test excitation can excite the jth column component, the element a of the corresponding position in the empty matrix ij When the ith row test excitation can not excite the jth column element, the element a of the corresponding position in the initial matrix is recorded as 1 ij And marking as 0, and obtaining a matrix formed by the to-be-selected test excitation and the element, wherein the initial matrix is 0-1.
3. The greedy algorithm based test excitation set optimization method according to claim 2, wherein: in the second step, x test excitations are sequentially selected from m test excitations to be selected by a greedy algorithm according to the initial matrix obtained in the first step 1 ，T 2 ，T 3 ...T x A set of test excitations F, F = (T) is formed 1 ，T 2 ，T 3 ...T x ) Obtaining the minimum cost sum S corresponding to the test excitation set F min The specific process comprises the following steps:
step two, firstly: the excitation of the component by the test excitation is represented by an n-dimensional boolean vector P, P if the jth column has been excited j True, otherwise P j = false, and the average cost c of each of m lines is calculated using equation (1) separately 1 ,c 2 ,...,c i ,...,c m
Step two: selecting the test excitation with the minimum average cost calculation result to excite the not excited components according to the calculation result of the average cost of each row in the second step, wherein the test excitation with the minimum average cost calculation result is the first element T in the selected test excitation set F 1 ；
Step two and step three: removing the excited components from the initial matrix to form a new matrix; re-executing the first step and the second step to select the second element T in the test excitation set F from the new matrix 2 Until all components in the circuit are excited, x test excitations are preferably selected to form a test excitation set F, F = (T) exciting the whole circuit 1 ，T 2 ，T 3 ...T x ) And the minimum cost sum corresponding to the test excitation set F is S min ＝s 1 +s 2 +...+s x 。
4. The greedy algorithm based test excitation set optimization method according to claim 3, wherein: the specific process of optimizing the test excitation set F in the second step by adopting an internal comparison strategy and a random elimination strategy in the third step is as follows:
step three, firstly: and optimizing the test excitation set F by adopting an internal comparison strategy, namely: test stimulus T in test stimulus set F w P components are excited, w is 1,2, \ 8230;, x, if T is divided by T in the test excitation set F w Other test stimuli can excite the same p components, so that T is converted into w Removing from the test excitation set F; if T is divided in the test excitation set F w If the other test stimuli cannot excite the same p components, T cannot be converted to w Removing from the set; obtaining an optimized test excitation set F;
step two: optimized test obtained from the third step and the first stepAn excitation set F is optimized by adopting a random elimination strategy; namely: randomly rejecting any two test excitations from the x test excitations in the test excitation set F each time, wherein the rejection times are k times,forming a new test excitation set after each elimination and recording the new test excitation set as F new The sum of the costs is S new (ii) a Removing any two test excitations from the optimized test excitation set F, generating q components which are not excited, executing the step two, selecting the test excitations for exciting the q components from the remaining m-x test excitations to be selected, and forming a new test excitation set F by the selected test excitations and the test excitations which are not removed from the test excitation set F new Calculating a new set of test stimuli F new Cost and S of new ；
Step three: new set of test stimuli F new Cost and S of new Cost sum S with set F min Comparison, if S min <S new If the set F is the optimal test excitation set; if S is min >S 1 Then use the new set F new Replacing F to be an optimal test excitation set after the first test excitation elimination; re-executing the third step and the second step until the rejection frequency k reachesSub or F new And obtaining an optimal test excitation set in the same way as the F.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710725698.7A CN107478981B (en) | 2017-08-22 | 2017-08-22 | Greedy algorithm based test excitation set optimization method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710725698.7A CN107478981B (en) | 2017-08-22 | 2017-08-22 | Greedy algorithm based test excitation set optimization method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107478981A true CN107478981A (en) | 2017-12-15 |
| CN107478981B CN107478981B (en) | 2019-12-10 |
Family
ID=60602121
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710725698.7A Active CN107478981B (en) | 2017-08-22 | 2017-08-22 | Greedy algorithm based test excitation set optimization method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107478981B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111398777A (en) * | 2020-03-10 | 2020-07-10 | 哈尔滨工业大学 | Simulation circuit test excitation optimization method based on synthetic deviation |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009125413A1 (en) * | 2008-04-09 | 2009-10-15 | Rafael Advanced Defense Systems Ltd. | Method and apparatus for gas detection based on spectral spatial misregistration |
| CN103576137A (en) * | 2013-09-27 | 2014-02-12 | 电子科技大学 | Multi-sensor multi-target location method based on imaging strategies |
| CN103986544A (en) * | 2014-05-07 | 2014-08-13 | 北京理工大学 | Spreading code cross-correlation optimization method of satellite navigation system based on greedy algorithm |
| CN104267294A (en) * | 2014-10-17 | 2015-01-07 | 东南大学 | IGIS fault online diagnostic method based on rough set theory and Petri network |
| CN106793120A (en) * | 2016-12-23 | 2017-05-31 | 西安电子科技大学 | Fast joint resource allocation methods based on partition number in virtual MIMO system |
-
2017
- 2017-08-22 CN CN201710725698.7A patent/CN107478981B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009125413A1 (en) * | 2008-04-09 | 2009-10-15 | Rafael Advanced Defense Systems Ltd. | Method and apparatus for gas detection based on spectral spatial misregistration |
| CN103576137A (en) * | 2013-09-27 | 2014-02-12 | 电子科技大学 | Multi-sensor multi-target location method based on imaging strategies |
| CN103986544A (en) * | 2014-05-07 | 2014-08-13 | 北京理工大学 | Spreading code cross-correlation optimization method of satellite navigation system based on greedy algorithm |
| CN104267294A (en) * | 2014-10-17 | 2015-01-07 | 东南大学 | IGIS fault online diagnostic method based on rough set theory and Petri network |
| CN106793120A (en) * | 2016-12-23 | 2017-05-31 | 西安电子科技大学 | Fast joint resource allocation methods based on partition number in virtual MIMO system |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111398777A (en) * | 2020-03-10 | 2020-07-10 | 哈尔滨工业大学 | Simulation circuit test excitation optimization method based on synthetic deviation |
| CN111398777B (en) * | 2020-03-10 | 2022-03-15 | 哈尔滨工业大学 | Simulation circuit test excitation optimization method based on synthetic deviation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107478981B (en) | 2019-12-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102821465B (en) | WLAN (Wireless Local Area Network) indoor positioning method based on subregion information entropy gain | |
| CN107943874B (en) | Knowledge mapping processing method, device, computer equipment and storage medium | |
| CN108734289A (en) | Method and apparatus for improving the robustness relative to " to resisting sample " | |
| CN110929867B (en) | Neural network structure evaluation and determination method, device and storage medium | |
| CN106291233B (en) | A kind of fault phase-selecting method based on convolutional neural networks | |
| CN113238227B (en) | Improved least square phase unwrapping method and system combined with deep learning | |
| CN101295008A (en) | Multi-target fault testing optimization method based on discrete particle swarm algorithm | |
| CN107133476B (en) | Test excitation and test point collaborative optimization method based on response aliasing measurement | |
| CN109583594A (en) | Deep learning training method, device, equipment and readable storage medium storing program for executing | |
| CN107478981B (en) | Greedy algorithm based test excitation set optimization method | |
| CN110221957A (en) | A kind of self-adapting random test method divided equally based on iteration region with positioning | |
| CN114445641A (en) | Training method, training device and training network of image recognition model | |
| CN110492468B (en) | False data attack identification method based on probability distribution | |
| CN108111188A (en) | Self-interference cancellation signal generation apparatus and method for full-duplex communication apparatus | |
| CN105096328A (en) | Camera self-calibration method based on IOS-PSO (Internetwork Operating System-Particle Swarm Optimization) | |
| CN106095692A (en) | The one-parameter sequence of events test data generating method of embedded software | |
| CN105188035B (en) | Indoor WLAN augmentation manifold positioned in alignment method based on the mapping of transition probability hot spot | |
| CN106934141B (en) | Acceleration method for calculating resistance based on long-edge cutting | |
| CN104809666B (en) | It is a kind of to search for the controllable small interference characteristic value calculating method of power system of radius of circle | |
| CN107608333B (en) | A kind of diagnosticability appraisal procedure based on equivalent depression of order | |
| Showkatbakhsh et al. | Secure system identification | |
| Xu et al. | A quasi-best random testing | |
| CN114896471B (en) | Chessman change detection method, chessman change detection device, computer equipment and storage medium | |
| CN117062002B (en) | 5G NR indoor positioning method and system based on lightweight TRANSFORMER | |
| CN112015610B (en) | Test sequence generation method, device, equipment and medium |
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 |