This paper first proposes a novel Multi-Universe Parallel Quantum Genetic Algorithm (MPQGA) and proves its global convergence in theory.In the algorithm,all individuals are divided into some independent sub-colonies,called universes,according to their definite topological structure.Individuals in a universe are represented by multi-state gene qubits.In the individual's updating,the general quantum rotation gate strategy and dynamic adjusting rotation angle mechanism are applied to accelerate convergence.Quantum NOT gate is used to realize quantum mutation to avoid premature convergence.Each universe evolving independently enlarges the search space.Information among the universes is exchanged by adopting the best emigration and the quantum crossover operation for the improvement of search efficiency.Then it puts forward a new Blind Source Separation (BSS) method based on the combination of MPQGA and Independent Component Analysis (ICA).The simulation results show that the efficiency of the new BSS method is obviously higher than that of the Conventional Genetic Algorithm (CGA) and the Quantum Genetic Algorithm (QGA).
This paper first proposes a novel Multi-Universe Parallel Quantum Genetic Algorithm (MPQGA) and proves its global convergence in theory.In the algorithm,all individuals are divided into some independent sub-colonies,called universes,according to their definite topological structure.Individuals in a universe are represented by multi-state gene qubits.In the individual's updating,the general quantum rotation gate strategy and dynamic adjusting rotation angle mechanism are applied to accelerate convergence.Quantum NOT gate is used to realize quantum mutation to avoid premature convergence.Each universe evolving independently enlarges the search space.Information among the universes is exchanged by adopting the best emigration and the quantum crossover operation for the improvement of search efficiency.Then it puts forward a new Blind Source Separation (BSS) method based on the combination of MPQGA and Independent Component Analysis (ICA).The simulation results show that the efficiency of the new BSS method is obviously higher than that of the Conventional Genetic Algorithm (CGA) and the Quantum Genetic Algorithm (QGA).
The genetic algorithm is an efficient optimization tool for its independence of problems, intrinsic parallelism and inherent learning capacity; however, it has some disadvantages such as slow convergence and prematurity. In this paper, a novel algorithm, called the quantum genetic algorithm—QGA, is proposed based on the combination of the quantum theory with the evolutionary theory. By adopting the qubit chromosome as a representation, QGA can represent a linear superposition of solutions due to its probabilistic representation. The random observation simulating the quantum collapse can bring diverse individuals, and the evolution of quantum chromosome can also pilot the evolution. So it has better diversity than the classical genetic algorithm. Rapid convergence and good global search capacity characterize the performance of QGA. The paper not only proves the global convergence of the QGA, but also some simulation experiments are shown to prove its superiority over other algorithms.
In order to improve the efficiency of the quantum genetic algorithm,this paper proposed a quantum genetic algorithm based on a expanded multi-chain coding scheme.The algorithm took qubit as chromosome.Each chromosome generated multiple and parallel gene chains which were mapping to multiple optimized solutions by separating qubit into multiple and parallel genes.The expanded genes chains expanded the searching space effectively and increased evolutionary rate for quantum genetic algorithm.It introduced the dynamic adjusting rotation angle mechanism to quantum rotation gate to guide individual evolution and used quantum not-gate to prevent algorithm occurring premature convergence.The method further improved searching efficiency.In the simulation experiment,analysed the influence for the algorithm with different variation probability([0,0.1,…,0.9,1])and used different code schemes to optimize extremal function.The simulation experiment result shows that it can obviously improve the efficiency of quantum genetic algorithm by adding gene chain,and the quantum genetic algorithm based on a expanded multi-chain coding scheme is efficient.
In this paper, a novel Quantum Genetic Algorithm (QGA) called Parallel Probability-gate Quantum Genetic Algorithm (PPQGA) is proposed.The main points of PPQGA are that a new chromosome representation called qubit representation, a new updating gate called probability-gate and a novel evolutionary strategy being used in update procedure are introduced.Based on the concepts and principles of quantum computing introduced, PPQGA is characterized by rapid convergence, good global search capability and the ability of possessing exploitation simultaneously.The experimental results of several special functions show that PPQGA can speedup the migration of the top individuals of subpopulations and is superior to other several genetic algorithms greatly in quality and efficiency.
In order to save computation time and avoid falling into local optima in quantum genetic algorithm (QGA), this paper proposes a quantum genetic algorithm based on Boltzmann selection mechanism (B_QGA). The proposed algorithm has several novel characteristics. Firstly, Quantum bits are encoded by angle and introduce the constant factor, aiming at enhancing calculated amount. Secondly, a new dynamic rotation angle adjustment strategy is presented to achieve updated chromosome, and individuals directly evolve to the current optimal solution with linear increasing rate, which greatly improve the rate of convergence and the ability of global search. Thirdly, instead of roulette selection mode, Boltzmann selection mechanism is employed for avoiding the local optimal problem. Finally, for the selected offspring's individuals, Hadamard door is proposed to perform mutation operation to sustain stability and diversity of population. The experiment results on three complicated continuous functions show the proposed algorithm is able to find the best solution in very short time and has excellent global search ability comparing to quantum genetic algorithm (QGA) and genetic algorithm (GA).
Multi-user detection based on quantum optimization algorithm has aroused intense interests recently. Quantum clone genetic algorithm has many shortcomings such as: low efficiency, poor population diversity, slow convergence speed, easy to trap in local minimums, blindness in global optimal searching direction and so on. An improved quantum clone genetic algorithm (IQCGA) is proposed in this paper. The algorithm have many merits: (1) Adopting nicking probability division initialization strategy, the population diversity is enhanced; (2) Introducing quantum whole interference crossover, the information is spread in the whole population, so it helps to avoid trapping in local minimums, and accelerates convergence speed; (3) Making use of adaptation strategy in quantum rotate gate updating, the searching speed for optimal solution is accelerated; (4) In order to avoid premature and evolution stagnation, the superior individual whole crossover quantum catastrophe strategy is adopted, which helps the population to search the objective solution in different directions. Simulation results prove that the multiuser detection based on given algorithm has lower bit error, rapider converge speed, better resisting MAI ability and better Near-far Resistance ability than other multi-user detections based on quantum clone genetic algorithm and classical genetic algorithms obviously.
The current quantum evolution algorithms have slow convergence rate and poor robustness. In order to overcome the two shortages, a novel self-adaptive quantum genetic algorithm is proposed. Firstly, the new algorithm adopts an encoding method which is based on the Bloch spherical coordinates. Secondly, in the process of searching the optimal solution, a self-adaptive factor is introduced to reflect the relative change rates which are relative to the difference of the best individual's objective fitness between the parent generation and the child generation. The convergence rate and direction of the algorithm can be improved by adjusting the factor. The rules of updating the rotation angle and are constructed. Finally, using hadamard gate of the quantum in the mutation strategy, it can enhance the diversity of population. The simulation results of the optimizing problem of the multidimensional complex functions show that the new algorithm has not only avoided effectively the premature and improved the convergence rate, but also boosted strikingly efficiency and stability robustness of the algorithm.