The conventional Frequency Diverse Array (FDA) system, using linear frequency offsets, forms periodic grating lobes and coupling, which may increase interference for potential users and difficulty in controlling parameters. To mitigate these issues and generate a dot-shaped beampattern, nonlinear frequency offsets are introduced to decouple the interdependence of parameters in the range and angle dimensions. Furthermore, to obtain the optimal nonlinear frequency offsets, we propose an algorithm based on the FDA Multiple-Input Multiple-output (FDA-MIMO) structure that integrates the Revised Dingo Optimization Algorithm (RDOA) with a Kaiser window function (referred to as the RDOAK algorithm) to form a dot-shaped beampattern. Specifically, the RDOA is used to optimize the nonlinear frequency offset coefficients, and the Kaiser window function is applied to adjust the waveform. Simulation results demonstrate the superior performance of our proposed RDOAK approach in preventing the mainlobe shift of the beampattern, eliminating grating lobes, suppressing jammings, and achieving narrower mainlobe width in the range dimension compared to other widely used algorithms.