7results about How to "Strong randomness" patented technology

The present application relates to the technical field of image generation, and particularly relates to an image generation training method for an AIGC large model, comprising: obtaining standard images by preprocessing collected labeled images, injecting the standard images into an initial model for training to obtain a pre-training diffusion model, inputting random noise into the pre-training diffusion model, and generating initial images by reverse diffusion; optimizing the pre-training diffusion model according to the initial images to obtain an AIGC large model with image generation capability, processing image generation requirements by using the AIGC large model to generate target images; determining whether the number of reverse diffusion sampling steps needs to be increased; determining whether the sampling temperature of the image generation process needs to be increased; and determining the residual connection weight of deep features based on the attack iteration decay rate in the optimization process. The present application improves the generation stability of the target images.

The invention relates to the technical field of wireless communication, and provides a spread spectrum sequence generation method and device, equipment, a medium and a product, and the method comprises the steps: generating a basic sequence, and generating a password sequence or a chaos sequence; performing field extension mapping on the basic sequence to form a base sequence, and performing field extension mapping on the password sequence or the chaos sequence to form a field extension sequence; performing extension field addition operation on the base sequence and the extension field sequence to obtain an extension field addition sequence; and carrying out extension domain transformation on the extension domain addition sequence to obtain a spread spectrum sequence. According to the method, the spread spectrum sequence meeting the requirement can be generated, and the method has good correlation, long periodicity, good randomness and high safety.

A method for planning stair-climbing torque based on multi-pose information fusion first acquires the pose information of the lower limbs when climbing stairs through a pose sensor. Then, based on the pose information of the lower limbs, the motion information of the human lower limbs when climbing stairs is obtained. Next, kinematic calculations are performed on the entire lower limb pose data to obtain the motion information of the human lower limbs, and then the human motion trajectory is obtained. The vertical velocity of the ankle joint is used as the discrimination criterion. When the velocity is greater than a predetermined threshold, an assist phase torque is output and the device outputs torque to follow the human body according to the interactive force data. When the velocity is less than the predetermined threshold, a follow-up phase torque is used and no assist phase torque is output. Finally, the motion controller MCU outputs torque and the knee joint motor executes it. This invention patent utilizes the ankle joint, which has good periodicity and strong noise resistance, to recognize the motion state and improve the stability of state identification. By filtering the human-computer interaction force and the output torque, the accuracy of torque is improved while optimizing the user comfort of the exoskeleton wearer.

The application provides a Doppler positioning method based on frequency hopping spread spectrum, and belongs to the technical field of radio positioning, to solve the problem that the fixed carrier frequency mode adopted by the traditional Doppler positioning technology makes the signal extremely easy to be intercepted, oriented and interfered. The method comprises the following steps: step 1: describing an application scenario; step 2: in the application scenario, a frequency hopping measurement signal is sent by a transmitting end; step 3: in the application scenario, a receiving end performs frequency hopping synchronization and measures a Doppler shift; step 4: a Doppler shift is preprocessed; and step 5: the preprocessed Doppler shift is used to calculate the position of a terminal to be positioned in combination with the relative motion of the transmitting end and the receiving end. The method can obtain the low interception rate and anti-interference ability given by the rapid change of the carrier frequency, and can also use the frequency diversity characteristics of frequency hopping to reduce the positioning performance deterioration caused by the multipath effect, and is suitable for application scenarios with high requirements for robustness, positioning accuracy and other performances.

The application belongs to the technical field of information processing, and particularly relates to a document-level relation extraction method based on a generative model, which comprises the following steps: firstly, in the training stage, a generative model is used to compress and refine the document, and the relation summary of the subject and object in the document is extracted, so that information closely related to the target is screened out; secondly, in the testing stage, a lightweight discriminant module is introduced, and the relation summary output by the generative model is mapped to a predefined relation category; and the generative model has self-adaptive learning ability and can dynamically and autonomously learn human carefully selected examples to generate output closer to human intention.

This invention relates to the field of chaotic system theory and discloses a quantum image encryption algorithm based on chaotic systems and DNA dynamic coding. The algorithm includes a quantum image coding module, a chaotic sequence generation module, a DNA dynamic coding and operation module, and a central control module. The algorithm also integrates a key management unit. By using an enhanced quantum phase representation method to convert classical images into quantum images, and utilizing quantum gate operations controlled by chaotic sequences to achieve quantum scrambling of pixel positions and preliminary diffusion of pixel values, the system can efficiently perform preliminary confusion and perturbation of image information directly at the quantum level, laying a secure preprocessing foundation for subsequent encryption. By using a parallel multi-chaotic system to generate a dynamic key stream and introducing a dynamic perturbation mechanism based on image hash values, the generated key stream has stronger randomness and is dynamically related to the plaintext image content, enhancing its ability to resist security threats.

This invention relates to a horizontal sand mill and its sand milling process for paint production, belonging to the field of grinding equipment technology. It includes a rotating reference barrel, a rotating power system, a sand milling dispersion system, and several adaptive actuating devices. The sand milling dispersion system is installed inside the rotating reference barrel, and the rotating power system drives the sand milling dispersion system. The adaptive actuating devices include an actuating reference ball, an upper sealing plate of the reference ball, a lower sealing plate of the reference ball, an actuating rotating shaft, an actuating rack, and a return spring. The actuating reference ball is installed on the sand milling dispersion system. The actuating reference ball has an actuating space arranged perpendicular to the rotation axis of the sand milling dispersion system. The upper and lower sealing plates of the reference ball are installed at both ends of the actuating space. The two ends of the actuating rack are respectively installed on the upper and lower sealing plates of the reference ball via return springs. The actuating rotating shaft and the actuating rack mesh and drive each other. The end of the actuating rotating shaft is equipped with an actuating plate, achieving sufficient flow of the grinding media and good grinding effect.