6results about How to "Stable on the ground" patented technology

A kind of double-runner dehumidification energy-saving optimization control method for multivariate collaborative regulation, comprising: obtaining the state observation data and equipment limit value of each measuring point, generating material regeneration demand estimation curve and desorption boundary threshold set.According to material regeneration demand estimation curve and desorption boundary threshold set, calculate the regeneration heat trajectory and residence time trajectory required for desorption, generate a set of safe operation constraints.Based on regeneration heat trajectory, residence time trajectory and a set of safe operation constraints, combined with the decision quantity corresponding to the set optimization target, determine the control setting sequence and reference path at the current time.The control setting sequence is issued to the execution unit, and the execution unit is scheduled according to the reference path, and the actual execution record and execution deviation data are monitored and output in real time.Based on actual execution record and execution deviation data, update the estimated value of material adsorption capacity and desorption temperature, and correct the regeneration heat trajectory and residence time trajectory in adjacent rolling time window.

The utility model relates to the technical field of soft roadbeds, in particular to a municipal highway engineering soft roadbed reinforcement treatment device. Comprising a bottom plate, a material conveying mechanism, a carding mechanism, a rolling mechanism and a moving mechanism, the upper half portion of the material conveying mechanism is arranged at the top of the bottom plate, the lower half portion of the material conveying mechanism is arranged at the bottom of the bottom plate, the rolling mechanism is arranged at the bottom of the bottom plate and arranged on the side close to a user, and the moving mechanism is arranged at the bottom of the bottom plate. The rolling mechanism comprises a roller, a rotating shaft and a fixed seal, vertical plates are fixedly arranged on the side, close to a user, of the bottom of the bottom plate, first through holes are formed in the vertical plates, the rotating shaft penetrates through the first through holes and is arranged between the two vertical plates, the two exposed ends of the rotating shaft are fixedly connected with the fixed seal, and the roller is fixedly arranged on the peripheral side of the rotating shaft. The rollers are driven by the rotating shafts to rotate on the bottom surface of the roadbed, so that the gravels are pressed into the roadbed, and the ground of the roadbed is more stable.

This invention discloses a method for the synergistic dezincification and desulfurization of high-sulfur bauxite, comprising the following steps: A) High-sulfur bauxite is crushed and fed into a ball mill, quicklime and circulating mother liquor are added according to the ratio for grinding to obtain a pretreated slurry; B) The pretreated slurry is fed into a reactor for leaching to obtain a leached slurry; C) The leached slurry is diluted, and after sedimentation and separation, a crude sodium aluminate solution is obtained; The molar concentrations of residual Zn²⁺ and S²⁻ in the crude sodium aluminate solution are detected, and the crude sodium aluminate solution is fed into a reaction tank and heated to 95-100℃. At the same time, lime milk is added to the reaction tank under stirring, and then stirring is continued at 95-100℃ for 20-40 minutes to obtain a precipitate; D) The precipitate is filtered to separate the precipitate and the filtrate; the filtrate is the crude sodium aluminate solution after dezincification and desulfurization. The method of this invention can achieve simultaneous and efficient removal of zinc impurities and S²⁻, reduce processing costs, reduce equipment corrosion, and improve the quality of alumina products.

The invention discloses a smart classroom emotion recognition method and system based on dynamic facial expressions. The method comprises the following steps: firstly, acquiring a classroom panoramic video stream, and preprocessing the classroom panoramic video stream to generate a structured facial image sequence; the method comprises the following steps of: obtaining a fused video frame through a space-time attention fusion module, then embedding time and space tokens, inputting the fused video frame into a double-scanning bidirectional Mamaba block to carry out space-time feature extraction and modeling, finally outputting the probability of each emotion through a pre-trained emotion classification model, and selecting the emotion with the highest probability as final output to be fed back to a user. The dynamic facial expression recognition system provided by the invention is light in weight and high in performance, so that facial expressions of students can be accurately recognized in an actual classroom, and the reliability and the stability of results are ensured. The method solves the problems of weak dynamic modeling, strong multi-mode dependence, lack of education closed loop and the like in a classroom scene in the prior art, and is suitable for non-sentible emotion perception and teaching optimization in intelligent education.

The invention discloses a kitchen raw and cooked food material cross contamination early warning system and method based on multi-modal fusion, and relates to the technical field of computer vision, and the method comprises the steps: obtaining a visible light video, a thermal infrared video and depth / binocular data in a kitchen operation region, completing the time-space registration, determining a target candidate of a kitchen operation entity, and carrying out the early warning of the cross contamination of the kitchen raw and cooked food materials; generating a multi-modal object candidate set; and performing relation judgment on the multi-mode object candidate set according to an open vocabulary video relation mode recognition rule, and outputting a triple event stream and a corresponding object identity track table. According to the method, an open vocabulary video relation mode is adopted to recognize and output a triple event stream and an object identity track table, so that interaction relations such as contact, bearing, transmission, taking and placement stably fall to the ground in an event segment form, and therefore, sequence division and pollution candidate link retrieval in a subsequent stage are supported.

The present application relates to the technical field of optical fiber sensing and machine learning, in particular to a plastic optical fiber automobile foot mat forming process parameter control system and method, aiming to solve the problem of insufficient parameter and real-time forming state adaptability caused by traditional control of non-fusion warm and pressure dynamic timing characteristics, leading to foot mat size deviation, shrinkage mark and poor optical fiber adhesion, the present application collects warm and pressure distribution data in the forming cavity in real time through the optical fiber sensing network, generates a standardized time series through normalization and time window segmentation; the pre-trained long short-term memory neural network model analyzes the timing dependency relationship through three hidden layers, generates a feature vector and maps it to the warm and pressure and pressure holding time set value, the execution control unit accurately controls the temperature through PID and pulse width modulation, dynamically calibrates the closed-loop pressure control through the proportional valve and pump speed, improves the parameter adaptability and ensures the stability of the forming quality.