Search

To address the issue of suboptimal spectral purity in Direct Digital Frequency Synthesis (DDFS) within resource-constrained environments, this paper proposes an optimized DDFS technique based on cubic Hermite interpolation. Initially, a DDFS hardware architecture is implemented on a Field-Programmable Gate Array (FPGA); subsequently, essential interpolation parameters are extracted by combining the derivative relations of sine and cosine functions with a dual-port Read-Only Memory (ROM) structure using the cubic Hermite interpolation method to reconstruct high-fidelity target waveforms. This approach effectively mitigates spurious issues caused by amplitude quantization during the DDFS digitalization process while reducing data node storage units. Moreover, this paper introduces single-quadrant ROM compression technology to further diminish the required storage space. Experimental results indicate that, compared to traditional DDFS methods, the optimization scheme proposed in this work achieves a ROM resource compression ratio of 1792:1 and a 14-bit output Spurious-Free Dynamic Range (SFDR) of −88.134 dBc, effectively enhancing amplitude quantization precision and significantly lowering spurious levels. This significantly improves amplitude quantization precision and reduces spurious levels. The proposed scheme demonstrates notable advantages in both spectral performance and resource utilization efficiency, making it highly suitable for resource-constrained embedded systems and high-performance applications such as radar and communication systems. [ABSTRACT FROM AUTHOR]

In order to solve the problem of the low end positioning accuracy of large hydraulic rock drilling robotic arms due to machining error and the working environment, this paper proposes an end positioning error compensation method based on an Improved Secretary Bird Optimization Algorithm (ISBOA) optimized Back Propagation (BP) neural network. Firstly, the good point set strategy is used to initialize the secretary bird population position to make the initial population distribution more uniform and accelerate the convergence speed of the algorithm. Then, the ISBOA is used to optimize the initial weights and biases of the BP neural network, which effectively overcomes the defect of the BP neural network falling into a local optimum. Finally, by establishing the mapping relationship between the joint value of the robot arm and the end positioning error, the error compensation is realized to improve the positioning accuracy of the rock drilling robot arm. The experimental results show that the average positioning error of the rock drilling robotic arm is reduced from 187.972 mm to 28.317 mm, and the positioning accuracy is improved by 84.94%, which meets the engineering requirements. [ABSTRACT FROM AUTHOR]