A Family of Asymmetrical Orthogonal Arrays with Run Sizes 4p^2

Nowadays orthogonal arrays play important roles in statistics, computer science, coding theory and cryptography. The usual difference matrices are essential for the construction of many mixed orthogonal arrays. But there are also many orthogonal arrays, especially mixed-level or asymmetrical which can not be obtained by the usual difference matrices. In order to construct these asymmetrical orthogonal arrays, a class of special matrices, so-called generalized difference matrices, were discovered by Zhang（1989, 1990, 1993） by the orthogonal decompositions of projective matrices. In this article, an interesting equivalent relationship between the orthogonal arrays and the generalized difference matrices is presented. As an application, a family of orthogonal arrays of run sizes 4p2, such as L36（6^13^42^10）, are constructed.

Orthogonal arrays obtained by generalized difference matrices with g levels

Nowadays orthogonal arrays play important roles in statistics, computer science, coding theory and cryptography. The usual difference matrices are essential for the construction for many mixed orthogonal arrays. But there are also orthogonal arrays which cannot be obtained by the usual difference matrices, such as mixed orthogonal arrays of run size 60. In order to construct these mixed orthogonal arrays, a class of special so-called generalized difference matrices were discovered by Zhang (1989,1990,1993,2006) from the orthogonal decompositions of projection matrices. In this article, an interesting equivalent relationship between orthogonal arrays and the generalized difference matrices is presented and proved. As an application, a lot of new orthogonal arrays of run size 60 have been constructed.

A SIMPLE METHOD FOR CONSTRUCTING ORTHOGONAL ARRAYS BY THE KRONECKER SUM

In this article, we propose a new general approach to constructing asymmetrical orthogonal arrays, namely the Kronecker sum. It is interesting since a lot of new mixed-level orthogonal arrays can be obtained by this method.

Parametric Optimization to Minimise the Surface Roughness on the Machining of GFRP Composites

The present investigation focuses on the parametric influence of machining parameters on the surface finish obtained in turning of glass fiber reinforced polymer （GFRP） composites. The experiments were conducted based on Taguchi＇s experimental design technique. Response surface methodology and analysis of variance （ANOVA） were used to evaluate the composite machining process to perform the optimization. The results revealed that the feed rate was main influencing parameter on the surface roughness. The surface roughness increased with increasing the feed rate but decreased with increasing the cutting speed. Among the other parameters, depth of cut was more insensitive. The predicted values and measured values were fairly close to each other, which indicates that the developed model can be effectively used to predict the surface roughness on the machining of GFRP composites with 95% confidence intervals. Using such model could remarkablely save the time and cost.

A Construction of Variable Strength Covering Arrays

Covering arrays(CA)of strength t,mixed level or fixed level,have been applied to software testing to aim for a minimum coverage of all t-way interactions among components.The size of CA increases with the increase of strength interaction t,which increase the cost of software testing.However,it is quite often that some certain components have strong interactions,while others may have fewer or none.Hence,a better way to test software system is to identify the subsets of components which are involved in stronger interactions and apply high strength interaction testing only on these subsets.For this,in 2003,the notion of variable strength covering arrays was proposed by Cohen et al.to satisfy the need to vary the size of t in an individual test suite.In this paper,an effective deterministic construction of variable strength covering arrays is presented.Based on the construction,some series of variable strength covering arrays are then obtained,which are all optimal in the sense of their sizes.In the procedure,two classes of new difference matrices of strength 3 are also mentioned.

Aerial vehicle guidance based on passive machine learning technique

Purpose-The purpose of this paper is to develop a new guidance scheme for aerial vehicles based on artificial intelligence.The new guidance scheme must be able to intercept maneuvering targets with higher probability and precision compared to existing algorithms.Design/methodology/approach-A simulation setup of the aerial vehicle guidance problem is developed.A model-based machine learning technique known as Q-learning is used to develop a new guidance scheme.Several simulation experiments are conducted to train the new guidance scheme.Orthogonal arrays are used to define the training experiments to achieve faster convergence.A wellknown guidance scheme known as proportional navigation guidance(PNG)is used as a base model for training.The new guidance scheme is compared for performance against standard guidance schemes like PNG and augmented proportional navigation guidance schemes in presence of sensor noise and computational delays.Findings-A new guidance scheme for aerial vehicles is developed using Q-learning technique.This new guidance scheme has better miss distances and probability of intercept compared to standard guidance schemes.Research limitations/implications-The research uses simulation models to develop the new guidance scheme.The new guidance scheme is also evaluated in the simulation environment.The new guidance scheme performs better than standard existing guidance schemes.Practical implications-The new guidance scheme can be used in various aerial guidance applications to reach a dynamically moving target in three-dimensional space.Originality/value-The research paper proposes a completely new guidance scheme based on Q-learning whose performance is better than standard guidance schemes.