A series of square-root-raised-cosine(SRRC) FIR filter with CSD coefficients were designed according to the local search algorithm based upon MiniMax error criteria. The simulation results of a baseband system show th...
详细信息
ISBN:
(纸本)0780381165
A series of square-root-raised-cosine(SRRC) FIR filter with CSD coefficients were designed according to the local search algorithm based upon MiniMax error criteria. The simulation results of a baseband system show that two 13-tap SRRC FIR filters with a roll-off factor 0.6 only introduced about 6% peak distortion in the eye pattern. A bit-level pipeline architecture was used to realize the high sampling rate FIR filter. An additional tap with fixed input of 10 was added to the final stage of the filter to avoid carry ripple. Consequently, the critical path consists of only a single one-bit full adder and a pipeline register. The filter was implemented in an altera's FPGA: EP20K60EFC144-1 and the timing analyses results show that the sampling rate could be over 200MHz.
Due to the simplicity of the Artificial Bee Colony (ABC) algorithm, it has been applied to solve a large number of problems. ABC is a stochastic algorithm and it generates trial solutions with random moves, however it...
详细信息
ISBN:
(纸本)9783319134611;9783319134604
Due to the simplicity of the Artificial Bee Colony (ABC) algorithm, it has been applied to solve a large number of problems. ABC is a stochastic algorithm and it generates trial solutions with random moves, however it suffers from slow convergence. In order to accelerate the convergence of the ABC algorithm, we proposed a new hybrid algorithm, which is called Memetic Artificial Bee Colony for Integer Programming (MABCIP). The proposed algorithm is a hybrid algorithm between the ABC algorithm and a Random Walk with Direction Exploitation (RWDE) as a localsearch method. MABCIP is tested on 7 benchmark functions and compared with 4 particle swarm optimization algorithms. The numerical results demonstrate that MABCIP is an efficient and robust algorithm.
The nested partitions method (NPM) is a global optimization method, which can be applied to solve many large-scale discrete optimization problems. The basic procedure of this method for solving the traveling salesman ...
详细信息
ISBN:
(数字)9783642549243
ISBN:
(纸本)9783642549243;9783642549236
The nested partitions method (NPM) is a global optimization method, which can be applied to solve many large-scale discrete optimization problems. The basic procedure of this method for solving the traveling salesman problem (TSP) was introduced. Based on the analysis and determination of the strategy of the four arithmetic operators of NPM, an improved NPM was proposed. The initial most promising region was improved by weighted sampling method;The historical optimal solution of every region was recorded in a global array;the 3-opt algorithm was combined in the localsearch for improving the quality of solution for every subregion;the improved Lin-Kernighan algorithm was used in the search for improving the quality of solution for surrounding region. Some experimental results of TSPLIB (TSP Library) show that the proposed improved NPM can find solutions of high quality efficiently when applied to the TSP.
Unlike a single-label supervisor dataset where each instance is assigned to one class label, in multi-label datasets, several class labels are assigned to each instance, which makes it difficult to build an accurate a...
详细信息
ISBN:
(纸本)9781665412414
Unlike a single-label supervisor dataset where each instance is assigned to one class label, in multi-label datasets, several class labels are assigned to each instance, which makes it difficult to build an accurate and comprehensive model from this dataset. In this study, a memetic algorithm for feature selection in a multi-label dataset is proposed. The principal innovation of this study is the offer of a novel local search algorithm which, in collaboration with binary quantum-inspired gravitational searchalgorithm (BQIGSA), forms the main framework of the proposed memetic algorithm. The main invention of the proposed local search algorithm is to build a number of neighbors for a solution using the prior knowledge vector and the posterior knowledge vector to select effective features and remove useless and irrelevant features. The results of implementing the proposed algorithm and comparing these results with similar works show that the proposed method in most cases leads to better results.
The Boolean satisfiability problem, abbreviated as SAT, is the backbone of many applications in VLSI design automation and verification. Over the years, many SAT solvers, both complete and incomplete, have been develo...
详细信息
The Boolean satisfiability problem, abbreviated as SAT, is the backbone of many applications in VLSI design automation and verification. Over the years, many SAT solvers, both complete and incomplete, have been developed. Complete solvers are usually based on the DPLL (Davis–Putnam–Logemann–Loveland) algorithm, which is a backtracking algorithm. Industrial strength problems are very large and make DPLL based solvers impractical for some applications. In such cases, local search algorithms that try to find a solution within a stipulated time can be used. These algorithms look at SAT problem as an optimization problem. They start with an initial random solution and explore a certain search space by iteratively making local changes to the solution using a greedy, heuristic algorithm to find a global optimum. Over the past few years, heterogeneous devices such as Graphics Processing Units (GPU) and Field Programmable Gate Arrays (FPGA) have been used to accelerate the SAT problem and handle large SAT instances. There has been a growing interest in exploiting the parallel and pipeline processing power of FPGAs for various applications. New process technologies have allowed for more logic blocks, memory elements, and faster FPGAs, making it a perfect candidate for parallel computing. This thesis presents a local search algorithm Walksat, implemented on the Xilinx Alveo U250 Accelerator card. The entire solver has been developed using the OpenCL framework. On-chip memory available on the FPGA has been exploited to a great extent and the solver can handle SAT problems of up to 98,000 variables and 401,800 clauses. We have also analyzed the performance of our solver against the state of the art complete and incomplete solvers.
In the scheduling literature, it is generally assumed that jobs are not split into sub-lots, or that the number and size of sub-lots are limited or predetermined. These assumptions make the problem more manageable. Ho...
详细信息
In the scheduling literature, it is generally assumed that jobs are not split into sub-lots, or that the number and size of sub-lots are limited or predetermined. These assumptions make the problem more manageable. However, they may prevent more successful schedules. For many businesses, considering the splitting of jobs while scheduling them can create significant improvement opportunities. This study addresses the Flexible Job-Shop Scheduling Problem (FJSP) with job-splitting, determining how many sub-lots each job should be split into and the size of each sub-lot. A MIP model is proposed for the considered problem. In the model, the size and number of sub-lots of a job are not predefined or bounded. The objective function of the model is to minimize the makespan. Feasible solutions could not be found for large-sized problems by the mathematical model. So, a Hybrid Genetic algorithm (HGA) is also proposed. In the proposed HGA, a local search algorithm (LSA) that determines the size of sub-lots has been included in the GA to improve the efficiency. To show the success of the proposed HGA, its performance is compared with the classical GA.
Background: Existing hidden Markov model decoding algorithms do not focus on approximately identifying the sequence feature boundaries. Results: We give a set of algorithms to compute the conditional probability of al...
详细信息
Background: Existing hidden Markov model decoding algorithms do not focus on approximately identifying the sequence feature boundaries. Results: We give a set of algorithms to compute the conditional probability of all labellings "near" a reference labelling lambda for a sequence y for a variety of definitions of "near". In addition, we give optimization algorithms to find the best labelling for a sequence in the robust sense of having all of its feature boundaries nearly correct. Natural problems in this domain are NP-hard to optimize. For membrane proteins, our algorithms find the approximate topology of such proteins with comparable success to existing programs, while being substantially more accurate in estimating the positions of transmembrane helix boundaries. Conclusion: More robust HMM decoding may allow for better analysis of sequence features, in reasonable runtimes.
With the increase in production levels, a pattern of industrial production has shifted from a single factory to multiple factories, resulting in a distributed production model. The distributed flowshop scheduling prob...
详细信息
With the increase in production levels, a pattern of industrial production has shifted from a single factory to multiple factories, resulting in a distributed production model. The distributed flowshop scheduling problem (DPFSP) is of great research significance as a frequent pattern in real production activities. In this paper, according to real-world scenarios, we have added blocking constraints and sequence-dependent setup times (SDST) to the DFSP and proposed a distributed blocking flowshop scheduling problem with sequence-dependent setup times (DBFSP_SDST). In a distributed environment, the allocation of resources and utilization have become an urgent problem to be solved. In addition, scheduling problems related to resource conservation have also attracted increasing attention. Therefore, we study DBFSP_SDST and consider minimizing the energy consumption cost of the critical factory (critical factory is the factory with maximum energy consumption cost) under resource balance. To tackle this problem, an effective iterated greedy algorithm based on a learning-based variable neighborhood searchalgorithm (VNIG) is proposed. In VNIG, an efficient construction heuristic is well designed. Two different localsearches based on the characteristics of the proposed problem are developed to enhance the local exploitation by neighborhood searching. A learning-based selection variable neighborhood search strategy is designed to avoid the solution trapping in local optima. By conducting extensive simulation experiments, the proposed VNIG shows superior performance compared with artificial chemical reaction optimization (CRO, 2017), the discrete artificial bee colony algorithm (DABC, 2018), the iterative greedy algorithm with a variable neighborhood search scheme (IGR, 2021), and the evolution strategy approach (ES, 2022).(c) 2022 Elsevier B.V. All rights reserved.
It is assumed that all machines will be used in studies dealing with parallel machine scheduling problems. However, for some businesses having special processes, where large furnaces with very intense energy consumpti...
详细信息
It is assumed that all machines will be used in studies dealing with parallel machine scheduling problems. However, for some businesses having special processes, where large furnaces with very intense energy consumption are used during commissioning, it can be very critical to complete jobs using the least number of furnaces. In addition, for many businesses, doing their jobs with fewer machines creates opportunities for unused machines to be rented to another company or to accept additional jobs as much as the capacity of idle machines. For this reason, in this study, the assumption that all machines will be used has been removed and a mathematical model has been proposed that will decide both which machines will be used and which jobs will be produced in which order on these machines, for the unrelated parallel machine scheduling problem with sequence and machine dependent setup times and machine eligibility restriction. The objectives of the considered problem are minimizing the number of machines to be used and the completion time of the last job. The objective functions of the proposed multi-objective mathematical model are scalarized using the weighted sum method. In order to show the solution performance of the mathematical model, randomly generated test problems were solved with GAMS / CPLEX. To solve the large problems, a local search algorithm and a genetic algorithm have been proposed due to the lack of feasible solutions with GAMS / CPLEX. In the large-scale problem, when all weight pairs are taken into account, genetic algorithm is more successful than local search algorithm an average of 25.64% in terms of solution quality and 50.31% in terms of time.
For two finite disjoint sets P and Q of strings over an alphabet SIGMA, an alphabet indexing psi for P,Q by an indexing alphabet GAMMA with \GAMMA\ GAMMA satisfying psi(P) and psi(Q) = empty set, where psi: SIGMA* --...
详细信息
For two finite disjoint sets P and Q of strings over an alphabet SIGMA, an alphabet indexing psi for P,Q by an indexing alphabet GAMMA with \GAMMA\ < \SIGMA\ is a mapping psi: SIGMA --> GAMMA satisfying psi(P) and psi(Q) = empty set, where psi: SIGMA* --> GAMMA* is the homomorphism derived from psi. We defined this notion through experiments of knowledge acquisition from amino acid sequences of proteins by learning algorithms. This paper analyzes the complexity of finding an alphabet indexing. We first show that the problem is NP-complete. Then we give a local search algorithm for this problem and show a result on PLS-completeness.
暂无评论