11results about How to "Improve polymerization effect" patented technology

The invention discloses a federated undersampling learning method for unbalanced network intrusion detection, which comprises a federated learning server and a client, and communication connection is established between the federated learning server and the client; all models for network intrusion detection are configured in the server; the server is used for maintaining a global model and deploying the global model to the client for training and learning, and the client adopts a local data set to train a local model downloaded from the server and uploads the local model after training and learning to the server; the server aggregates the local models uploaded by the client to obtain a global model with updated parameters; and when the client adopts the local data set to train the server, undersampling processing is performed on the local data set to obtain a balanced data set to perform learning training on the client. When federated learning is carried out on an intrusion detection network model, the influence of training data imbalance in a client on the model detection performance is solved through an undersampling method.

The invention discloses a federal learning-based human body behavior recognition method and system and a storage medium, and relates to the field of human body behavior recognition. The method comprises the following specific steps: acquiring a human body behavior recognition data set; preprocessing the human body behavior recognition data set to obtain a first data set; performing various data imbalance characteristic analysis on the first data set; designing a federated learning aggregation algorithm FedPA, and dynamically allocating weights for client model aggregation according to the sum of client model parameter quantities; and dividing the first data set into a training set and a test set, training the algorithm FedPA by using the training set, and evaluating the trained algorithm FedPA by using the test set. According to the method, the human behavior recognition accuracy under the federated learning framework is effectively improved, the aggregation performance of the server and the client model is optimized, the data privacy problem is relieved, and the communication overhead is reduced.

The invention relates to the technical field of organic chemistry, in particular to a preparation method of low-viscosity glycidyl acrylate, which comprises the following steps: 1, reacting hydroxyethyl acrylate with succinic anhydride under the action of a catalyst to obtain a compound 2; 2, the compound 2 obtained in the step 1 reacts with epoxy chloropropane to enable epoxy to be subjected to ring opening, and a compound 3 is obtained. And 3, adding sodium hydroxide into the reaction liquid of the ring-opening product compound 3 in the step 2 at room temperature, heating, and carrying out ring closing to obtain a target product 4. Compared with some existing photopolymerizable bifunctional epoxy resin, the epoxy resin has the advantage of low viscosity. The production process is mild and simple, has relatively low requirements on equipment, and is suitable for industrial production.

The invention discloses a slump-retaining water reducer mother liquor and a preparation method thereof, and the slump-retaining water reducer comprises the following raw materials: alkylene alkenyl polyoxyethylene ether, hydrogen peroxide, an acrylate polymerization initiator mixture, an acrylate copolymerization initiation solution, a lipid-containing gel substance, a sodium hydroxide solution and deionized water. The lipid-containing gel substance can enhance the water-retaining property of the water reducer and reduce the water loss of concrete in the stirring and transportation processes, so that the stability of the concrete is improved, sedimentation and layering are prevented, the lipid-containing gel can react with free chloride ions in water to generate a water-insoluble gel substance, the permeability of the concrete is reduced, and the water-retaining property of the concrete is improved. The water reducing agent is favorable for improving the durability of concrete and reducing permeation of moisture and harmful chemical substances, so that the service life of the concrete is prolonged, the water reducing efficiency of the water reducing agent is improved, good workability of the concrete is allowed to be kept under the condition of lower dosage, meanwhile, the flowability and plasticity of the concrete are improved, and pouring and forming are facilitated.

The application discloses an organic silicon emulsion, a preparation method and a method for dehydrating and shaping water-saturated wooden cultural relics, and relates to the field of water-saturated wooden cultural relic protection research.The preparation method of the organic silicon emulsion comprises the following steps: taking 2-5 parts of organic silicon monomer, 1-3 parts of a surfactant, 0.3-1 parts of a co-surfactant and 3-6 parts of pure water for standby; the surfactant and the co-surfactant are added into the organic silicon monomer, and uniform stirring or oscillation is continuously carried out to form a transparent organic silicon dispersion liquid; and the organic silicon dispersion liquid is poured into the pure water, and sufficient stirring or oscillation is carried out to form a stable organic silicon emulsion.The organic silicon emulsion material prepared by the application has good dehydrating and shaping performance, and can significantly enhance the dimensional stability, mechanical properties, microbial resistance and environmental stability of the cultural relics, and is environment-friendly and efficient.

The present invention discloses a robot scene recognition method based on lifelong learning, which improves the robot's scene recognition accuracy and lifelong learning ability by integrating advanced feature extraction technology and aggregation technology, combining long-term and short-term memory mechanisms, and intelligent knowledge distillation and memory-aware synaptic strategies. The lifelong learning robot scene recognition method of the present invention has significant performance improvements, especially in complex and changing environments. Compared with the existing technology, the method can provide higher recognition accuracy and better long-term learning ability. Through application and testing on Nordland datasets in multiple seasons and Oxford datasets in multiple lighting conditions, the present invention demonstrates its powerful recognition ability for different scenes, as well as its ability to balance the retention of old knowledge and the learning of new knowledge in lifelong learning scenarios.

The invention relates to a preparation method of a carbodiimide cross-linking agent. The preparation method comprises the following steps: under the protection of nitrogen, putting diisocyanate into a 500mL four-mouth flask, and stirring; the temperature is raised to 180-230 DEG C, an initiator and a catalyst are added, the initiator is triisopropyl phosphite, the catalyst is triphenyl bismuth, and the content of isocyanate is measured after reaction; and when the content of the isocyanate is reduced to be less than 10%, cooling, and discharging while hot to obtain a polycondensation product polycarbodiimide. Compared with the prior art, the preparation method has the advantages that diisocyanate is taken as a reaction raw material, the specific initiator and the catalyst are added, so that the reaction can be carried out more quickly, the polymerization effect is better, the synthesis operation is simple, and the obtained product has the characteristics of low toxicity, water solubility, quick age hardening, enough retention time and the like; the polycarbodiimide provides an effective way for the development of a high-performance water-soluble coating agent.

An episulfide composition, including: an episulfide compound (A) that comprises an episulfide ring, and a polysulfide bond in which three or more sulfide bonds represented by - S- are linked together; and an episulfide compound (B) that comprises an episulfide ring and a (di)sulfide bond but not the polysulfide bond, wherein the content of the episulfide compound (A) is 100 ppm by mass or more with respect to a total amount of the episulfide composition.

The invention belongs to the technical field of high polymer material synthesis, and particularly discloses a method for preparing semi-aromatic nylon by dispersion polymerization of a non-polar solvent, which comprises the following steps: carrying out salt forming reaction on aliphatic diamine and aromatic dibasic acid to prepare nylon salt; and carrying out polymerization reaction on the nylon salt in a non-polar solvent to prepare the semi-aromatic nylon. The method solves the problems of caking and kettle sticking in the production process of semi-aromatic nylon, and is easy to discharge; the polymerization reaction is completed in one step, the reaction condition is mild, the reaction time is short, and the product molecular weight is high.

One embodiment according to the present disclosure relates to a monomer for an optical component, including: a compound represented by the formula (1), in which a pH obtained through a pH measurement method below is 4.0 to 7.5. (pH Measurement method) 80 g of a sample, 80 mL of toluene, and 250 mL of water are mixed together in a liquid separation funnel, an aqueous phase is separated off, and the pH of the aqueous phase is measured.

The present application belongs to the technical field of distributed machine learning and edge intelligence, and discloses a self-adaptive quantization decentralized learning method and system for heterogeneous edge devices. The present application is directed to a plurality of edge devices with heterogeneous computing resources, heterogeneous storage capabilities and time-varying communication links. Under the condition of no central server participation, through training parameter and communication relationship initialization, training quantization scale self-adaptive determination, low-precision local training, communication quantization scale self-adaptive determination, quantization model information exchange and neighbor aggregation update, decentralized collaborative learning is completed. The present application can simultaneously alleviate the resource limitation problem of heterogeneous edge devices in the training stage and the communication stage, ensure the implementability of decentralized collaborative learning, improve resource utilization efficiency and improve model convergence stability.