58 results about "Immune network" patented technology

To provide a new method for autonomously controlling the behavior of a control target device based on a stimulating action and a suppressing action among antibodies in an immune network, an operating unit 3 calculates an antibody concentration ai(t) serving as an index for selecting an antibody module ABi, while plural antibody modules ABi different in stimulating conditions are set as processing targets. A convergence judging unit 4 judges whether the antibody concentration ai(t) is converged to a predetermined target value ri. When a judgment of non-convergence is made, a convergence controlling unit 5 calculates a correction parameter ul(t) for correcting the antibody concentration so that the antibody concentration ai(t) approaches to the target value ri. When a judgment of convergence is made, an antibody estimating unit 7 calculates an estimation value Pi, and selects some antibody module ABi based on the estimation values Pi calculated for the plural antibody modules. The behavior of the control target device is controlled in accordance with a control content defined by the selected antibody module ABi.

To provide a new method for autonomously controlling the behavior of a control target device based on a stimulating action and a suppressing action among antibodies in an immune network, an operating unit 3 calculates an antibody concentration ai(t) serving as an index for selecting an antibody module ABi, while plural antibody modules ABi different in stimulating conditions are set as processing targets. A convergence judging unit 4 judges whether the antibody concentration ai(t) is converged to a predetermined target value ri. When a judgment of non-convergence is made, a convergence controlling unit 5 calculates a correction parameter ul(t) for correcting the antibody concentration so that the antibody concentration ai(t) approaches to the target value ri. When a judgment of convergence is made, an antibody estimating unit 7 calculates an estimation value Pi, and selects some antibody module ABi based on the estimation values Pi calculated for the plural antibody modules. The behavior of the control target device is controlled in accordance with a control content defined by the selected antibody module ABi.

The invention provides a traditional Chinese medicine composition for treating the acute exacerbation of bronchial asthma. The traditional Chinese medicine composition is subjected to prescription according to monarch, minister, adjuvant and conductant, wherein achyranthes root and earthworm together serve as monarch drugs, cicada slough, ossa draconis, oyster, turtle shell and white peony root together serve as ministerial drugs, thunberbg fritillary bulb, amethyst, figwort root, radix asparagi, rhizoma anemarrhenae, turtle shell, white peony root, toosendan fruit and raw malt serve as adjuvant drugs, and licorice serves as a conductant drug. The traditional Chinese medicine composition fully prepared according to the prescription has the efficacies of endopathic wind calming, asthma relieving and wheeze stopping. Shown by modern medicine, the bronchial asthma is a typical psychosomatic disease, and the liver occupies an important position in the psychosomatic medicine of the traditional Chinese medicine and plays a crucial role in a nerve-endocrine-immune network, so that the treatment for the bronchial asthma from liver not only is in accordance with the theory of traditional Chinese medicine, but also has the scientific basis of the modern medicine, and is one of the important methods for treating the asthma. The traditional Chinese medicine composition provided by the invention has the advantages that the excessive and contrary liver Yang can be excellently calmed, the strong and contrary liver wind can be relieved, and the acute contraction of a lung tube is slowly spread, so that the acute exacerbation of the asthma is quickly eased, and no adverse reaction occurs; and the effective percentage is 80%.

The invention is applicable to the field of internet information safety and provides an immunological network system. The system comprises a transparent firewall, an intelligent inspection device and an emergency device, wherein the transparent firewall is used for analyzing and extracting the scan features and preventing external network scanning; the intelligent inspection device is used for monitoring the flow which enters the network, extracting the attach fingerprint features according to the abnormal flow and storing the attach fingerprint features into an immunological feature library; and the emergency device provides an emergency channel for the damaged internal network nodes detected by the intelligent inspection device and prompts the user to reduce the damaged nodes to the safety state before attack after the operation of the user is finished. The network flow is monitored and audited, the good state of the network is maintained, the unknown intrusion behavior is analyzed and memorized, the network immunity is improved, the damage range can be effectively controlled after intrusion, the unimpeded network service can be normally provided, the system has the self-restoring and reducing capacity, and the stable operation of the network is comprehensively maintained.

The invention discloses an improved immune network abnormal behavior detection method, including four stages of autologous database data extraction, antigen presentation, abnormal behavior detection and clone selection, characterized in that: the autologous database data extraction, antigen presentation Both adopt a single-category self-data generation model based on a deep belief network; the deep belief network is stacked by a restricted Boltzmann machine (RBM for short); the RBM is a neural network; the abnormal behavior detection A method based on the combination of innate immunity and adaptive immunity is adopted; the clonal selection adopts a method of clonal variation based on generative network. The present invention introduces the model and method of deep learning into the computer immune network anomaly detection model, improves the quality and efficiency of model training, greatly improves both detection efficiency and accuracy, and overcomes the randomness of traditional methods. Or the defect that cross-mutation leads to a large number of invalid calculations.