195results about How to "Reduce overdependence" patented technology

The invention provides a composite plant solid drink suitable for diabetic patients, hyperglycemia and fat crowds. The solid drink comprises the following components in parts by weight: 40-65 parts of grain puffing powder, 4-10 parts of medicine and food homology plant extract, 5-20 parts of functional ingredient, 0.1-0.5 part of compound vitamin and 2.5-4.5 parts of compound mineral. According to the composite plant solid drink, the 'medicine and food homology' of the traditional Chinese medicine is combined, the modern nutriology and the biological hygiene are applied, the raw material ratio and the nutrition content meet the diet requirements of the diabetic patients, namely high dietary fibers, low fat, high-quality proteins, a proper amount of vitamins and minerals can meet the nutrition meal replacement and addition demands of the diabetic patients, the diabetic complication is prevented or relieved, the solid drink can be matched with drug for treatment, the glucose fluctuation of the diabetic patients is controlled, the over-dependence of the diabetic patients on drugs is reduced, and even the glucose tolerance of the diabetic patients can be recovered.

The invention discloses a routing method for a heterogeneous network based on cognition, mainly solving the problem that the existing method considers only part of environment factors of the network, and neglects the dynamic complexity of a heterogeneous network environment so as not to cognize and predict the environment. The method comprises the following steps: firstly, judging the types of the existing available networks, perceiving a network environment state, evaluating the network environment state by using a Q-learning method and calculating end-to-end time delay and the estimated values thereof; secondly, using a utility function based on multiple parameters to obtain the utility values of various networks according to the perceived network environment state; then selecting the network with the biggest utility value by a probability of P=0.9, and selecting other available networks as target networks by the probability of 1-P; and finally continuously transmitting businesses, implementing network switch and updating the network environment state by the target networks. The routing method improves the utilization ratio of network resources, achieves load balance of the network and can be used in the heterogeneous network environment.

The invention discloses a pulmonary nodule diagnosis method based on a local receptive field and semi-supervised depth self coding. Firstly, the local receptive field is adopted to carry out multilevel feature extraction on a pulmonary nodule image; then, semi-supervised sparse self coding is used to autonomously learn nodule features in a pulmonary image; and finally, multiple pieces of clinical information are fused to realize accurate detection on the pulmonary nodule. The recognition accuracy is high, the inputted CT image can be subjected to feature multiple classification and then self coding learning, the network diagnosis is more accurate, and an important role in auxiliary diagnosis is played for a doctor.

The invention discloses a method of realizing pulmonary nodule medical sign recognition based on image retrieval with semantic features and supervised Hashing. The method includes a step A of extracting pulmonary nodule mixed sign region in a lung CT image, and cutting off the same for each single sign regions; a step B of extracting semantic features expressing pulmonary nodule sign information using a convolutional neural network (CNN) based on parameter sharing; a step C for realizing similar pulmonary nodule image retrieval; and a step D for recognizing pulmonary nodule signs. The method of the invention is based on the pulmonary nodule image retrieval with the semantic features and supervised Hashing, the sign category shown in the pulmonary nodule image is thus recognized, a doctor is assisted in judging the benign and malignant degrees of pulmonary nodules, and over-reliance of the doctor on diagnostic experience is reduced.

The invention discloses a resource treatment method for household garbage and biochemical sludge. The resource treatment method for the household garbage and the biochemical sludge is characterized by comprising garbage sorting, sludge drying, low-temperature dry distilling, melting and burning, oil and gas separating and purifying and flue gas treating processes and comprising the following specific steps: (1) sending the household garbage to the garbage sorting process and sorting out useful fuel, and sending the biochemical sludge to the sludge drying process to obtain the dried sludge; (2) sending the useful fuel and the dried sludge to a low-temperature dry distilling process to obtain dry-distilled oily gas and dry-distilled residues; (3) sending the dry-distilled oily gas and the dry-distilled residues to the oil and gas separating and purifying, and melting and burning processes, separating out tar and glassy solid residues; and (4) treating flue gas generated in the melting and burning process by the flue gas treating process, and discharging the flue gas into upper air. The problem that dioxin is generated when the garbage and the sludge are burned can be solved easily; resource utilization of the household garbage and the biochemical sludge is facilitated; the garbage sorting difficulty is reduced; and the resource treatment method for the household garbage and the biochemical sludge is convenient in industrial application.

The invention discloses a GPU-based modeling method for automatic machine learning. The method comprises the steps of establishing a model library; preparing data; conducting the normalization treatment on the feature part of the input data; selecting a plurality of models in the common mode to subject the models to the GPU parallel computation; automatically preparing data for cross verification according to the input data, wherein the default is 10 folds: 9 folds of training and 1 fold of verification test; respectively calculating the cross verification for the plurality of models through the GPU to obtain the errors of all the models respectively; selecting one model smallest in error as a finally selected model; adopting all data as training data; training the finally selected model to obtain a final model, and marking the final model as Ypre = f (x). In this way, an optimized model is automatically calculated and obtained for a common user. Therefore, the complexity of the machine learning application for the common user and the excessive dependence of the common user on experts are reduced.

The invention provides a method for manufacturing a soybean protein modified adhesive formaldehyde-free medium-density fiberboard, and belongs to the technical field of density fiberboard manufacture. The method comprises the following steps: heating, stirring and mixing raw materials (soybean pulp, a denaturing agent, a crosslinking agent, a defoaming agent, a preservative and water) to react; preparing the soybean protein formaldehyde-free adhesive by virtue of a colloid mill and a homogenizer; adding a water-proofing agent, a curing agent and a buffering agent into the soybean protein formaldehyde-free adhesive, stirring and mixing at room temperature, and uniformly mixing with wood fibers; performing drying, paving and hot-pressing to prepare the soybean protein modified adhesive formaldehyde-free medium-density fiberboard. The method is environment-friendly and non-pollution; the soybean protein modified adhesive formaldehyde-free medium-density fiberboard is high in adhesion strength, good in water proofing, low in price, low in viscosity and easy to treat and process, can be subjected to hot-pressing and cold-pressing in application, and is not easy to crack. The soybean protein modified adhesive formaldehyde-free medium-density fiberboard can be adhered to fibers and the like, and the processing process can be controlled, and an environment-friendly fiberboard can be manufactured.

The invention discloses an achieving method for a branch identification and electric energy measuring type line loss analyzer. The analyzer solves the problem that line loss anomalous points of a zone area with the complex power grid structure are hard to position rapidly. The power frequency load carrier signal technology is adopted as the principle, and on the basis of the quasi-synchronous sampling core algorithm, the branch recognition and electric energy measuring synchronous operation functions are achieved. The analyzer is composed of a host and three current clamps, the host comprises a load carrier signal sampling processing module, an electric parameter collecting operation module, a load carrier signal generating protecting module, an embedded processor module, a display and touch screen control module, a communication module, an electric energy data and information storage module and other modules. The load carrier signal sampling processing module is subjected to load carrier signal demodulation and operation processing, and the branch topology affiliations are divided step by step. The electric parameter collecting operation module completes power line parameter sampling and operation.

The invention provides a construction method for a simplified genome next generation sequencing library based on double enzyme digestion and a kit. Aiming at defects of an existing construction method for the double enzyme digestion simplified genome next generation sequencing library, the double enzyme digestion combined range is expanded, and excessive dependence on expensive instruments of constructing the simplified genome library is reduced, the library construction flow path is simplified, library construction cost is reduced, the sequencing efficiency is improved, and meanwhile the technology is easy and flexible to operate and easier for researchers to master and can be realized in a common molecule lab. The construction method is particularly suitable for miniature or medium-scale labs needing to conduct SNP molecular marker development, genetic map construction, population genetics research, phylogeny biological research and the like on a great number of species with incomplete reference genomes. The construction method has good practical application value and application prospects in the fields of molecular breeding of agriculture, conservation biology and evolutionary biology.

The invention provides a TiO2 nanotube array photoanode and a preparation method thereof. The photoanode is characterized by electrodepositing a silver film on the surface of the TiO2 nanotube array of the photoanode, wherein the microscopic surface topography of the silver film is randomly distributed dendrites; the primary dendrites are 6-15mu m long; the secondary dendrites are 600nm-3mu m long; and the space between the secondary dendrites is 100-300nm. The silver modified TiO2 nanotube array photoanode is obtained by reasonably controlling the concentration, electrodeposition temperature, deposition potential and deposition time of the silver nitrate electrolyte. The silver is used as the noble metal instead of gold and electrodeposition is used instead of the electron beam lithography technology to prepare the silver nanoparticle-TiO2 nanotube array structure as the solar cell photoanode, thus lowering the cost and shortening the preparation period. The short-circuit current density and open-circuit voltage of the silver nanoparticle-TiO2 nanotube array photoanode after being modified by the metal are greatly improved compared with the short-circuit current density and open-circuit voltage of the silver nanoparticle-TiO2 nanotube array photoanode before being modified by the metal.

The invention discloses a system and a method for extracting a global LFP (Late Follicular Phase) Parkinsonism feature value. The system comprises a signal transceiving module, a feature extraction module, a comprehensive processing module and an output module, wherein the signal transceiving module is used for generating stimulation pulses and measuring LFP signals at the same time; the feature extraction module is used for performing energy analysis on the LFP signals measured by the signal transceiving module, and extracting the feature value of each band; the comprehensive processing module is used for calculating correlation coefficients of bands and energy baseline spectra according to the difference between the feature value of each band and each corresponding point on the energy baseline spectra; and the output module is used for adding the products of the feature values and correlation coefficients of the bands to obtain a global LFP Parkinsonism feature value. The LFP signals are preprocessed into a low frequency band, an intermediate frequency band and a high frequency band; the preprocessed signals are converted from a time domain to a frequency domain for analyzing, an autoregressive model and a neural network model are established using the feature value of each band as a parameter and using the correlation coefficient of each band as a weight, and a global feature value which is objective and accurate and does not depend on doctor experience is obtained.

The invention provides a filter tip pre-filled with large-particle materials. The filter tip is composed of a filter element, the large-particle materials, forming paper or groove-shaped cellulose paper. The forming paper wraps the filter element, and the large-particle materials are pre-buried in the filter element. The particle diameter of the large-particle materials ranges from 1.0 to 4.0 mm, and the distance between the buried position and the sucking end of the filter tip is greater than 6.0 mm. A flavour ingredient is contained in the large-particle material, or an absorbent material ingredient is contained in the large-particle materials. A layer of surface film may be arranged on the outer surface of the large-particle materials, and the surface film is broken due to the temperature of smoke in the case that the filter tip is sucked, so that a flavour is easy to emit, or an adsorption effect is easy to act. A filter stick pre-filled with large-particle materials can be formed in the case that a plurality of filter tips pre-filled with large-particle materials are integrally connected. The filter tip provided by the invention can add the flavour of a cigarette, or improve the adsorption capacity for smoke. The internal large-particle materials cannot be sucked in a mouth by mistake, and moreover, the flavour and the adsorption capacity of the large-particle materials are greatly preserved during production and storage, and can exert the corresponding effects during sucking.

The invention discloses a fractional order characteristic driven micro electro mechanical system (MEMS) control method. The control method comprises the following steps: under the condition of considering the input hysteresis characteristic, establishing a mathematical model of the MEMS, by considering the unmeasurable variable, time-varying lagging item, uncertainty and the output constraint thereof of the system, disclosing a chaotic oscillation phenomenon of the MEMS, and constructing time-varying delay in a compensation system state through Lyapunov-Krasovskii function, and proposing a self-adaptive stabilization control scheme for fusing an extension state observer, a Chebyshev neural network, a tangent baffier function and a tracking differentiator in a frame of the backstepping. Thecontrol method has the following effects of accelerating the solving speed, improving the system transient and steady-state performance, broadening the limitation of a physica sensor, reducing the computation workload, simplifying the controller design, cancelling the requirements on the primary function center and width, reducing the dependence on the precise model, improving the precision of atraditional first-order low-pass filter, overcoming an explosion problem of the differential item and guaranteeing that the output constraint condition cannot be violated.

The invention discloses a novel self reset shock insulation support seat, comprising an upper connection plate and a lower connection plate. The novel self reset shock insulation support seat is characterized in that corresponding grooves are formed in middles of the upper connection plate and the lower connection plate; bulges are correspondingly formed at the middles of the grooves; elastic cushion layers are cemented at the surfaces of the grooves; an oval rolling body is movably connected between the elastic cushion layers; two trenches are symmetrically formed in the spherical surface of the rolling body; a steel plate laminating rubber drum connected with the upper connection plate and the lower connection plate is arranged at the outside of the rolling body; a clearance is reserved between the rolling body and the steel plate laminating rubber drum; lead cores are arranged at the inner peripheral direction of the steel plate laminating rubber drum at equal intervals. The support seat can absorb high vibration energy, is high in self recovery capacity, and has the advantages of simple structure, convenience in use, high safety and the like, and excessive lateral displacement of the support seat is avoided without lifting a structure body.

The invention discloses a design method of a digitalized relief, and the method comprises the following steps: obtaining digitalized relief data through a laser scanning device, and extracting the relief from a background surface to directly finish an the preliminary modeling of the relief; and using an editing tool to modify the position and shape of the relief, and then integrating the relief with the background surface to realize the design process of the digitalized relief. By using the design method, the undue dependence on the manual skill is reduced, the labor cost is reduced, plenty of time is saved, and the design period of the digitalized relief is shortened. The design method has the advantage of repeatability in design.

The invention discloses a mechanical arm action learning method and system based on third-person imitation learning. The method and system are used for automatic control of a mechanical arm so that the mechanical arm can automatically learn how to complete a corresponding control task by watching a third-party demonstration. According to the method and system, samples exist in a video form, and the situation that a large number of sensors are needed to be used obtaining state information is avoided; an image difference method is used in a discriminator module so that the discriminator module can ignore the appearance and the environment background of a learning object, and then third-party demonstration data can be used for imitation learning; the sample acquisition cost is greatly reduced; a variational discriminator bottleneck is used in the discriminator module to restrain the discriminating accuracy of a discriminator on demonstration generated by the mechanical arm, and the training process of the discriminator module and a control strategy module is better balanced; and the demonstration action of a user can be quickly simulated, operation is simple and flexible, and the requirements for the environment and demonstrators are low.

The invention discloses a silicone-modified vegetable oil polymer polyether defoamer and a preparation method thereof, wherein the preparation method comprises the following steps: introducing a siloxane chain into a molecular chain of vegetable oil by utilizing the reaction of epoxy vegetable oil and silicone oil, to obtain siloxane-modified vegetable oil polyols; performing transesterification on siloxane-modified vegetable oil polyols in a methanol solution of potassium hydroxide to obtain siloxane-modified vegetable oil methyl ester; the ring-opening polymerization is performed on siloxane-modified vegetable oil methyl ester, ethylene oxide and propylene oxide to obtain the silicone-modified vegetable oil polymer polyether defoamer. According to the invention, the silicone-modified vegetable oil polymer polyether defoamer solves the problems of poor hydrophilicity and low foam breaking rate of vegetable oil defoamer, is stable in emulsion properties, good in water dispersion property, no non-toxic and free of physiological activity. New ideas for the synthesis of environment-friendly defoamers are provided, and more new options for food grade defoamers are provided.

The invention relates to a fault eliminating method based on evaluation of detecting points, which comprises the steps of: listing a fault tree according to a fault phenomenon; determining fault probability and detecting cost; establishing logic structure weight of each detecting point through dichotomy; comprehensively evaluating the detecting points by the three factors for determining a preferred detecting point; then, detecting the detecting point; comparing a detecting result with a theoretical value; eliminating a no-fault branch in an original fault tree by judging whether the detecting point is correct; updating the fault tree; evaluating each detecting point in a new fault tree again; accordingly, selecting a new preferred detecting point for detection; repeating the steps for cycling the preferred detection; and positioning fault until only an output detecting point of a certain fault point is wrong in the fault tree and input detecting points are all correct. According to the invention, factors of the fault probability of fault points, the detecting cost of the detecting points, the logic structure weight of the detecting points in the whole fault tree and the like are comprehensively considered; and the method is simple and easy, has strong comprehensiveness and can provide fault elimination work with a guiding basis.

The invention belongs to the technical field of drawing analysis, and discloses a method for dynamically arranging smoke detectors in a CAD drawing. The method comprises a dynamic arrangement method and a component analysis module, and the specific operation steps of dynamically arranging smoke detectors in a CAD drawing are as follows: S1, obtaining a building base drawing dwg file to be drawn; S2, through a cad analysis service, obtaining information of floors related to the drawing and information of the most basic primitives of the components. According to the invention, a deep neural network model is utilized to solve the problem of accurate recognition of components and spaces in CAD architectural design drawings, and excessive dependence of component recognition on layers is reduced; based on the identified components and spaces, the problem of arrangement of smoke detector components in electrical design is solved; by means of accurate calculation of a computer, it can be guaranteed that placement of the components completely conforms to design specifications, and compared with manual design and manual drawing, labor cost and time cost are saved to the maximum extent.