Research On Time Modulated Array Theory And Conformal Array Technology

With the rapid development of modern wireless communications and electronics technology, the electromagnetic problems are increasing complex, which will result in a extremely strong demand for the high efficiency optimization technology. However, the optimization objective functions in the electromagnetic problems are always nonlinear, non-differentiable and have multiple peaks, which severely restrict the application of the traditional optimization algorithms. As a type of global optimization algorithm, evolutionary algorithms are simple and convenient to use, have weak limitations on the optimization objective functions and are not restricted by the search space. Therefore, evolutionary algorithms are widely applied to all kinds of electromagnetic optimization problems. However, for the complex optimization problems, its convergence rate is slow and it is easy to trap into the local optimal solutions. Therefore, it is significance in both theory and application to research efficient evolutionary algorithms.In the pattern synthesis, traditional phased arrays usually require a high dynamic range of amplitude excitations, which raises a high requirement on the feed system of arrays. By introducing time as a new degree of freedom in the traditional phased array antennas, time modulation array can simplify the feed network effectively and achieve array feeding accurately. Therefore, it is a potential research work to develop novel design methods of time modulated arrays and exploit new applicationsSince conformal arrays can fit the surface of the carrier tightly, it can utilize the carrier platform efficiently and can keep the most original carrier platform of aerodynamic characteristics and greatly reduce the radar cross-section of the whole system. Besides, its beam can achieve a wide angle scanning. The unique advantages of conformal arrays can effectively make up for the planar arrays inherent defects and conformal arrays can also meet the demand for the development of modern radar and other electronic systems. Therefore, conformal arrays have received wide attention in recent years and will have a very wide application prospect in the aircraft, satellites and ships.This dissertation makes deeply insight into evolutionary algorithms, time modulated arrays and conformal arrays. The author’s major contributions can be summarized as follows:1. Aiming at overcoming the drawbacks of the standard evolutionary algorithms such as slow convergence speed and easily trapping into the local optimal solution when solving complex problems, a hybrid artificial bee colony-differential evolution algorithm is proposed, which inherits the good global optimization ability of differential evolution algorithm and retains the strong local optimization ability of artificial bee colony algorithm. The convergence rate of the evolution is greatly increased. The proposed algorithm can avoid trapping into local optimal solution. Therefore, the performance of the proposed algorithm is improved.2. Aiming at overcoming the drawback that large time modulated arrays is difficult to synthesized, a synthesis method for large time modulated arrays is proposed. The method firstly synthesizes the pattern at the center frequency preliminarily and obtains an initial solution. Then the amplitude and phase of the initial solution will be optimized by a perturbation method to meet the requirements of the center frequency pattern and equivalent amplitude excitation of the array. This method decomposes the original large-scale optimization problem, which can alleviate the problem that the optimization ability will drop sharply when the scale of the optimized variables is too large.3. Based on the research of the feature that traditional time modulated arrays can achieve equivalent amplitude weighting at the center frequency, phase modulated arrays are proposed. Each antenna of the phase modulated arrays is connected with different delay lines. By controlling the operation sequence of the two delay lines, controllable equivalent phase excitations at the center frequency are introduced. The equivalent phase weighting at the center frequency can be achieved. Therefore, beam steering can be realized without phase shifters. The structure of the arrays can be greatly simplified and the arrays have a wide application prospect.4. The synthesis of time modulated arrays by evolution algorithms will bring slow computation rate and low stability. And its optimization ability is sensitive to the scale of variable problems. Aiming at overcoming the drawbacks, a fast synthesis method for time modulated arrays is proposed. Firstly, the center frequency pattern is synthesized by alternative projection algorithm. Utilizing the optimized result, sideband frequency pattern is synthesized by alternative projection algorithm again. Finally, the static and dynamic excitations are separated. Since the proposed approach employs alternative projection algorithm, high calculation efficiency and high stability can be obtained and the proposed method can synthesis large scale time modulated arrays.5. Aiming at solving the issue in the polarization direction adjustment of conformal arrays, a polarization direction controlling method is proposed for dual port antenna elements. According to the locations of the antenna elements and the polarization direction of the main lobe, the current relative relationship between the two ports can be obtained. Based on the relationship, the current amplitudes of the two ports are determined. Therefore, the direction of the current on the radiating element surface can be adjusted and the polarization directions of elements can be controlled.6. Aiming at solving the issue in low sidelobe level pattern syntheses of conformal arrays, a novel low sidelobe synthesis method based on unimodal surface optimization is proposed. By utilizing the modified Bernstein polynomial, a cluster of spline curves is generated over the aperture. These spline curves are used to fit a controllable unimodal smooth surface, where the amplitude excitations of the array are sampled. Then, the optimization of the amplitude excitations is converted to the optimization of the unimodal surface fitted by the spline curves, which is able to reduce the numbers of optimization variables greatly. Moreover, by utilizing the unimodal characteristics of the amplitude excitation distribution, the optimization process can be greatly reduced.