1489 results about "Adaptive capacity" patented technology

The invention discloses an image reflection removing method based on a depth convolution generative adversarial network. The method comprises the following steps: 1) data obtaining; 2) data processing; 3) model structuring; 4) loss defining; 5) model training; and 6) model verifying. In combination with the capability of a depth convolution neural network to extract high-level image semantic information and the flexible capability of the generative adversarial network to define the loss function, the method overcomes the limitations of a traditional method using low level pixel information, therefore, making the method with a better adaptive ability to the reflection removal of a generalized image.

The invention relates to a biomimetic quadruped robot provided with head and tail balance adjustment devices. The biomimetic quadruped robot comprises a trunk, a head balance adjustment device, a tail balance adjustment device and four robot legs, wherein the four robot legs are distributed under the robot trunk, and the head balance adjustment device and the tail balance adjustment device are located on a trunk front portion cross beam and a trunk back portion cross beam. The biomimetic quadruped robot uses a digital steering engine as a driver, and the steering engine can achieve speed and position control and feed back speed, positions, force moment and other information and facilitates on-line observation of a motion state of the robot. Each leg is provided with three driving freedom degrees and one driven extension freedom degree, three-dimensional space motion can be achieved, and the robot can have strong adaptive capacity to complex environments and obstacle crossing ability. The centre-of-gravity position of the robot can be timely adjusted when the robot with the head and tail balance adjustment devices walks in a static state, and the stability of the robot is improved. When the robot is in a unstable state, namely turns over, and reverse rotation force moment is produced by rapidly rotating the head and the tail to prevent the robot from turning over and being restored to be in a balanced state. The biomimetic quadruped robot provided with the head and tail balance adjustment devices is suitable for the multiple fields of military affairs civil cargo transportation, anti-terrorist devices, filed probing, planetary exploration and the like under complex terrain environments.

The invention interacts with a subject to query, challenge and identify multiple aspects and multidimensional influences that provide access to, allow modulation of, and entrainment of, state-specific global bio-regulatory self-organizing controllers and evocable triggers. Modulation of state-specific regulatory triggers may evoke optimized newly emergent self-organizing principles within the subject and ultimately support up-regulation of “states of presence” that include newly emergent controllers of additional optimal regulation of bio-chemical expressions. The “state” measures of a person may include the status of a combination of any number of identified biological qualities. The invention includes systems and methods that support access to a subject's state controller functions, for various aliments, to empower shifting ones' biology from a symptomatic to an asymptomatic state and to optimal adaptive learning and readiness. The system of the invention investigates and accesses the capacities that control such state shifts so that they can be broadly challenged, expanded, and entrained for optimized global regulatory function in reversing a myriad of pathological symptoms, learning limitation, and adaptive dysregulations.

The invention discloses a variable-rigidity soft body driver, soft body arm and soft body platform based on a blocking principle. A vacuum pump and a unit module are included. The unit module comprises an inner flexible framework, gas pipes, an outer flexible layer, a blocking mechanism, a deformation plug and a blocking plug. The inner flexible framework comprises a center through hole and deformation holes evenly distributed in the circumferential direction. The blocking mechanism is fixed in the center through hole and comprises supporting frameworks, elastic connection ropes, large ball particles, small ball particles, tightening springs and a sealing thin film. The elastic connection ropes sequentially penetrate symmetry centers of all the supporting frameworks. A containing cavity isformed between every two adjacent supporting frameworks. The containing cavities are filled with the large ball particles and the small ball particles. The corresponding gas pipes are arranged in theblocking mechanism and each deformation hole correspondingly. The other end of each gas pipe is connected with the vacuum pump. The variable-rigidity soft body driver, the soft body arm and the softbody platform have excellent adaptive capacity to the environment, and by means of the variable-rigidity principle, large flexibility can be achieved, and large rigidity can also be achieved.

A route planning method for a mobile robot comprises the steps that the distance between the mobile robot and an obstacle is detected through a plurality of distance sensors; a fish swarm algorithm with an elimination mechanism is provided; the mobile robot is controlled by a controller with a neural network; the neural network updates a neural network weight through the fish swarm algorithm; the neural network is trained through a known map template; the walking route of the mobile robot in environmental space with unknown environment information is planned through the trained neural network. The controller with the neural network structure controls the mobile robot through the fish swarm algorithm with the elimination mechanism added, the robot only needs to be trained in one template map, the behaviors of obstacle avoiding and destination reaching can be learned through the generalization performance of the neural network, a generalized behavior is learned through the algorithm rather than a route track of a specific map, and the adaptive capacity of the robot for complex position environment is improved.

The invention provides a hexapod-robot real-time gait planning method based on deep reinforcement learning. The method comprises the following steps of using a hexapod robot to acquire environment road condition information and making an integral movement track; through a camera, acquiring an environment photograph, according to the photograph, using a binocular range finding method to calculate road condition information of a target track, and using the calculated road condition information of the track in robot center of mass movement track navigation; in a foot end swinging space range of robot legs, taking photographs of a road condition environment, and through a trained deep reinforcement learning network based on a depth determinacy policy gradient (DDPG), carrying out data dimension reduction and feature extraction on the photograph; and according to a feature extraction result, acquiring a control policy of a hexapod robot, wherein the hexapod robot controls foot laying of a robot according to the control policy so that real-time walking of the hexapod robot is realized. By using the gait planning method, a complex and non-structural environment of a road condition can be planned in real time. The method has an important meaning for increasing an environmental adaptive capacity of the hexapod robot.

The invention belongs to the field of internet of things and machine vision and discloses an urban expressway traffic real-time monitoring method based on an information physical network. The system realized the method comprises a video sensing unit, a video image processor and a network communication and traffic monitoring client. The system achieves accident detection based on a video transmission feedback control method through a road traffic parameter detection, traffic accident initial automatic early warning and traffic state quantization identification and by using of network communication, completes real-time monitoring on urban expressway traffic, improves real-time performance and an accuracy rate of an existing traffic accident detection method and adaptability on an emergent traffic accident, has stronger adaptive capacity on traffic states and environment, fully utilizes time and space information through a traffic abnormal predicting result fusing time and space on a decision-making level, and avoids traffic abnormal missing detection.

The invention provides a face recognition method of attitude robust. The method comprises the following steps of: detecting a face region of an input face image, locating a face key characteristic point and carrying out unitary processing on the face region; evaluating the attitude of an input face, and obtaining the class of the attitude of the input face; extracting the characteristic of the face region subjected to the unitary processing; obtaining a linear combination's coefficient by utilizing a ridge regression model according to the extracted characteristic of the face region and the face data coupled on the basis of the attitude class; and inputting the linear combination's coefficient into a classifier to obtain the result of face recognition. By applying the method, the sensibility of face recognition to attitude difference is lowered, and the adaptive capacity of face recognition to various application scenarios is enhanced while better recognition performance is kept.

The invention relates to a signal transmitting device for realizing wireless measurement while drilling. A downhole signal transmitting device for an electromagnetic measurement while drilling system comprises an upper drilling string and a lower drilling string, and an insulating joint is contained between the upper drilling string and the lower drilling string; and the upper drilling string and the lower drilling string are used for forming an asymmetric dipole antenna. The transmitting device adopts software radio technology to realize various functions of wireless transmission; the impedance matching between a transmitter antenna and transmission medium is realized through an impedance converter; and the dipole antenna is adopted to transmit a measurement while drilling data signal. The device and the method improve the adaptive capacity and measurement depth of the electromagnetic measurement while drilling system to different strata.

The invention discloses an intelligent cognitive wireless network system for realizing heterogeneous network convergence. New function entities with multi-domain active cognition, high-efficiency heterogeneous network resource management, and network or terminal reconstruction are added in the network elements of the existing wireless network system, namely, a cognitive information management module, an intelligent management module and a heterogeneous network convergence management module are added on a network side; a reconstruction management unit, a united wireless resource management & self-optimization unit, a cognitive information / knowledge base and cognitive information management module are added on a terminal side; four interactive interfaces, namely interfaces among operators, interfaces between the network side and the terminal side, interfaces among respective modules and interfaces among units inside the modules are arranged. Modules or units inside the modules realize information exchange via interfaces, and realize corresponding network functions jointly, so that the cognitive capacity, autonomous capacity and adaptive capacity of the network are extended, the intellectualization of network management is promoted, the efficient utilization of heterogeneous network convergence and system frequency spectrum is realized, and overall performance of the network is promoted.

The invention discloses a method, a system and a device for intrusion detection. The method comprises the following steps: acquiring (201) abnormal behavior characteristics of a message; extracting (202) primary abnormal intrusion regulations according to the abnormal behavior characteristics; searching (203) a subsequently received message in a matching way according to the primary abnormal intrusion regulations; considering (205) that intrusion behaviors exist if the subsequently received message is matched to be searched. The embodiment of the invention shortens the detection response time of abnormal intrusion and enhances the intrusion discovery rate and the adaptive capacity of the abnormal intrusion behaviors.

The invention discloses a machine vision based mushroom automatic picking up system and method. The mushroom automatic picking up system includes a camera, a main control system, a motion system, an air bag type mechanical arm, and an air pump. The problem that the efficiency of mushroom manual picking is low, the sorting quality is uneven, and the conventional picking end executors can cause damage to mushroom and are poor in deformation and adaptive capacity can be overcome; the system can process mushroom seedling bed images having complex backgrounds and serious adhesion, and the problem of damage to the mushroom due to picking-up can be solved by the air bag type mechanical arm. The machine vision based mushroom automatic picking up system and method can provide a new method for mushroom industrial and automatic production, can achieve full mushroom production automation from seedling bed management to picking and sorting, can reduce the production cost, and can improve the classification quality.

The invention relates to an identification method of a remote sensing image building, belonging to an identification method of a remote sensing target image; the method overcomes the problems of false detection and miss-detection of the building in the existing detecting method of the building for improving the accurate detecting rate. The identification method sequentially comprises the following steps: region division, initial extraction of the building, road filtration and following extraction of the building. The identification method has good stability and strong environment adaptive capacity, and can solve the problem that the buildings and the roads close to the adjacent optical spectrum are confused, and the method can detect the building target with different shape structure and optical spectrum characteristics in the same image, the extracting result of the building has high accurate rate and good robustness, and the method can effectively extract various building targets under the remote sensing images with high resolution, therefore, the identification method has practical value in the construction of a geographical information system and a digital city system.

The invention provides a preparing method of powder metallurgy tantalum powder and / or niobium powder. The method comprises the steps that (1) heat treatment sintering is conducted, wherein heat treatment sintering is conducted on raw tantalum powder and / or the niobium powder and after sintering is accomplished, cooling is conducted and then passivation is conducted so that a sintered block can be obtained; (2) hydrogenation and powder preparation are conducted, wherein the sintered block obtained through heat treatment is hydrogenated and powder is prepared so that the hydrogenated tantalum powder and / or the hydrogenated niobium powder can be obtained; (3) dehydrogenation and oxygen reduction are conducted, wherein reducibility metal is added to the hydrogenated tantalum powder and / or the niobium powder so that dehydrogenating and deoxygenating heat treatment can be conducted, and therefore a mixture of reducibility metallic oxide, residue reducibility metal, dehydrogenated and deoxidized tantalum powder and / or the niobium powder can be obtained; (4) acid pickling and impurity removal are conducted, wherein acid pickling and drying are conducted on a mixture obtained through dehydrogenation and oxygen reduction so that the tantalum powder and / or the niobium powder suitable for powder metallurgy can be obtained. The method is suitable for recycling and treating the waste tantalum powder and / or the niobium powder generated in the process of production of the high-specific-volume tantalum powder and / or the niobium powder for a capacitor so that the tantalum powder and / or the niobium powder suitable for powder metallurgy can be prepared, investment of needed devices is small, the adaptive capacity to raw materials is strong, the technological process is short, and the safety coefficient of the treatment process is high.

The invention relates to a method for calculating the expected car following distance in driver car following behavior analysis. On the basis that drivers design multiple different driving scenes during data collection on a car surrounding simulation testing platform, driving data is collected aiming at the drivers; the car following behavior feature parameter data set of each driver is extracted from the driving data; the feature behavior parameter data sets of all the drivers are clustered into several different driving classifications, and are adopted as training data sets; then the training data sets are utilized for carrying out classification judgment on the current drivers to be classified. Different car following behaviors are classified, the efficiency of collecting car following behavior data is improved, the cost is low, and good safety is achieved. The longitudinal driving behaviors of the drivers can be simulated through the Gussian process, the individualized expected car following distance can be provided for the current drivers, and the active adaptive capacity of an auxiliary driving system to different drivers is improved.

The invention discloses a road surface construction robot environment perception system and method based on a multi-source sensor. A camera and a laser radar are arranged on a robot body; an embedded industrial control computer is arranged in the robot body; the embedded industrial control computer is used for obtaining and fusing class information of picture data and distance information of the laser radar based on a data fusion multi-target detection ranging algorithm, then, tracking a target by adopting a particle filter algorithm in the two directions of an image and point cloud, obtaining dynamic estimation change of an obstacle target in continuous time, and finally, obtaining the advancing speed and the motion direction of an obstacle in a construction area through the dynamic estimation change; road surface environment perception on the construction area is realized by drawing a risk graph; therefore, the adaptive capacity of the system for different environments is obviously improved; furthermore, the shielding condition between obstacles has some improvements on multi-target detection and a tracking algorithm; therefore, the effective operation of the construction robot is ensured; and thus, the system and the method have good application prospect.

A low-power media access control method of node dynamic sleeping includes the following steps: (1) organizing the network topology to clustering topology and realize a coarse-grained clock synchronization between the node inside the cluster and the cluster head and make the node work and sleep periodically; (2) the nodes have three operating modes: a sending mode, a receiving mode and a cycle work mode, wherein the node of the cycle work mode is characterized in that the nodes work or sleep periodically and the actual work and sleep time period of the nodes changes dynamically and the operation sub-time-period of the neighboring node can be calculated; (3) in one work cycle, the operation sub-time-period of any two adjacent nodes can be overlapped to guarantee the communication; (4) the node can intercept the packet sent by the adjacent node and adjust the sleep time period according to the information carried by the packet to reduce the carrier interception times; (5) dynamically adjusting the busy and idle degree of the node according to different data sampling rates so as to reduce the power consumption while guaranteeing the throughput. The method, which is simple and reliable, has advantages of low power consumption, short delay, strong adaptive capacity and effectively prolonged life cycle of the network.

The invention discloses a detection system and method for wear defects of a track, belonging to the field of wear detection. The detection system and method for wear defects of the track have high measurement precision, high work efficiency and strong adaptive capacity to environments. The system is installed on a track inspection car and used for defect detection of tracks on a to-be-inspected section; a plurality of nodes are arranged on the to-be-inspected section, and each node corresponds to a standard three-dimensional model; three-dimensional reconstruction is carried out on the surface of the tracks by using a structured light sensor; a processing unit compares a three-dimensional point cloud obtained by the structured light sensor with a corresponding standard three-dimensional model, detection is carried out so as to obtain absolute wear loss of the tracks, and a defect area is marked; all the three-dimensional point clouds are spliced via the processing unit so as to obtain a complete three-dimensional cloud picture; so the system has strong adaptive capacity to environments. The detection method for wear defects of the track has the advantages of high measurement precision and low cost.

The invention discloses a barrier crossing method and a barrier crossing device, belonging to the technical field of automatic cleaning. The method is applied to a cleaning robot with two parallel driving wheels. The method comprises the following steps: when the cleaning robot moves, detecting that whether the cleaning robot is in the barrier obstructing state or not; if the cleaning robot is in the barrier obstructing state, acquiring the inclination angle of the cleaning robot; and if the inclination angle is smaller than a preset angle, controlling the cleaning robot to cross the barrier. With the barrier crossing method and the barrier crossing device, the problem that during the moving process, the cleaning robot can not continue the cleaning task due to the obstruction of the barrier, and only can continue the cleaning task with the assistance of the outside, is solved, so that the cleaning robot can interpedently cross the barrier, and the work adaptive capacity of the cleaning robot is improved.

The invention discloses a method for predicting the cycle life of a fused lithium ion battery based on an EKF (Extended Kalman Filter) method and an AR (AutoRegressive) model, namely the method for predicting the cycle life of the lithium ion battery. The purpose is to solve the problem that the current methods based on models have a low adaptive capacity to different batteries and different working states. The method comprises the following steps: 1, measuring the capability data of the lithium ion battery to be measured on line, storing the data, and preprocessing the data; 2, based on the EKF method, determining the parameters of the state space model of the lithium ion battery; 3, according to the established state space model of the lithium ion battery, estimating the state of the lithium ion battery to be measured, and utilizing the output of the AR model to update the state of the lithium ion battery to be measured; causing the state space model of the lithium ion battery to obtain the capability data of the battery in each charging and discharging cycle, and comparing the data with the failure threshold of the lithium ion battery to be measured to obtain the residual life of the lithium ion battery. The method is used for predicting the cycle life of the lithium ion battery.

The invention discloses a method for building a covered wire inner full-state turbofan engine vehicle-mounted real-time model. The method comprises the steps of firstly building a vehicle-mounted real-time dynamic model, indicating an applied force-included and above-idle-speed state, of a turbofan engine, then, building a starting simplified model of the turbofan engine, finally, utilizing a Kalman filter for estimating gas circuit performance characteristic parameters of the turbofan engine, and building a vehicle-mounted self-adaptation real-time model indicating the applied force-included and above-idle-speed state. The method can meet the requirement of a covered wire full-state general model of the turbofan engine for precision and vehicle-mounted real-time performance, the strong adaptive capacity for model mismatch caused by the engine individual difference and performance degradation is obtained, model foundation is provided for experimental verification of a modern control theory and a fault-tolerant control method, and the method plays a positive and promotion role in shortening the preparation time, reducing test risk and cost.

The invention discloses a four-foot bionic robot with a planar four-bar metamorphic mechanism used on the waist. The four-foot bionic robot comprises a closed chain and four open chains. The closed chain is a planar four-bar metamorphic bar mechanism composed of a first bar, a second bar, a third bar and a fourth bar with the same length. The open chains are of the same structure. Each open chain is a 2R serial mechanism and connected with one bar in the closed chain through a rotating hinge. Bars in each open chain are located in the same plane. The planar four-bar metamorphic mechanism is applied to the waist of the four-foot bionic robot; by means of the characteristic of variable structure stages and variable degrees of freedom of the metamorphic mechanism, waist deformation of the four-foot bionic robot is achieved, so that the waist structure of a four-foot animal is better simulated, simulation of various kinds of animals is achieved, the advantages of various kinds of animals are integrated, and flexibility and adaptive capacity of the four-foot bionic robot for different terrain environments are improved.

The invention relates to a potting technology for mangrove forest seedling cultivation, which is characterized by comprising the steps of 1) selecting a seedling cultivation field, wherein muddy inning intertidal zones with weak storm, low salinity and fresh water supply are selected; 2) cultivating the seedlings in a nutrition bag, wherein a seedling cultivation matrix is filled in the nutrition bag, and direct seedling cultivation with the nutrition bag is adopted for vivipary mangrove forest plants; 3) transferring the bagged seedlings into pots, wherein plantlet grows up after the seedling is cultivated for 0.5-1 year in bags according to the method of step 2, and then the plantlet is transplanted to a plastic seedling cultivation pot for seedling cultivation; 4) cultivating the potted seedling, wherein the seedling cultivation is carried out after the plantlet is transplanted into the plastic cultivation pot and arranged in a nursery garden, and water and fertilizer management, reshaping and trimming as well as pest control are strengthened in the seeding cultivation process; and 5) outplanting of the seedling, wherein the mangrove forest seedling is cultivated after the plantlet transplanted in step 4) grow for 1-2 years in the plastic seedling cultivation pot. At the time, the seedlings can be transferred out of the nursery garden to be used for recovery of the mangrove forest of the inshore intertidal zone and greening of the coastal intertidal zone. The seedlings cultivated by the technology disclosed by the invention has the characteristics of being high in degree of lignification of the stem, complete and dense in roots and the like; and compared with plantlets of the mangrove forest, the seedlings have stronger adaptive capacity to hard intertidal zones and severe growing environments.

The invention relates to a Bayer image compression method based on YUV conversion, which belongs to the technical field of an aerospace optical remote sensor and is a Bayer image compression method used for an aerospace optical remote sensing CMOS camera. The method comprises the following steps: carrying out space conversion on a Bayer-format image from a three-color space to a brightness-chroma space, eliminating the relativity among the three color components to be ready for different compression ratios used by brightness and chroma, self-adaptively determining the compression ratios based on the complexity of the image, carrying out line conversion and encoding. The invention greatly improves the quality of the compressed image by carrying out space conversion to eliminate the relativity among the three color components, and further enhances the adaptive capacity of the camera by using different compression ratios for images of different complexities.

The invention provides a capture spin elimination mechanism suitable for spin space fragments. The capture spin elimination mechanism comprises a rotating part and a fixed part. Multiple finger mechanisms with underactuation mechanism forms are arranged on a shell at the front end of the rotating part, each finger mechanism is provided with multiple rotating joints, each joint is provided with a torsion spring, each finger is dragged by one rope to form an underactuation mechanism, bending actions can be completed, and adaptive capacity is achieved for contact shapes. The torsion spring of each joint resists traction force of the corresponding rope, and unbending power is provided for the fingers. All the fingers are arranged pairwise in an intersecting mode, and objects of different sizes can be held when the fingers close up. The ropes of all the fingers are fixed to a sliding block, and linkage can be achieved. The sliding block is located in the rotating part, can do translational motion relative to the rotating part along the rotating axis, but cannot rotate relative to the rotating part. The sliding block is driven by a lead screw-nut mechanism, and the sliding block is connected with a nut through a rotating pair.

The invention discloses a six-redundancy-type optical fiber strapdown inertial navigation system (SINS). The system is composed of a structural component part, an inertial device part and a peripheral circuit part, wherein the inertial device part is installed on a regular dodecahedron IMU (inertial measurement unit) support of the structural component part, and the peripheral circuit part is distributed on the system body brink of the structural component part. In the invention, the regular dodecahedron is designed to be of a hollow structure, so that high stiffness and light weight of a system IMU table body are realized; redundant configuration is adopted, so that the accuracy and reliability of the inertial navigation system are improved; the separation of the optical path and circuit of an optical fiber gyroscope is basically realized in structure for a system optical fiber gyroscope, so that varying degrees of influences of temperature and temperature gradient change brought by circuit heating on the optical path are avoided; the system has stable performance and good adaptive capacity to environment, can bear a variety of environment test requirements similar to that born by the inertial navigation system, and has good real-time performance.

The invention provides a magnetorheological elastomer intelligent vibration absorber used for ships, which comprises an outer shell, a magnetorheological elastomer, a magnetic coil group and an internal iron core. The outer shell, the magnetorheological elastomer, the magnetic coil group and the internal iron core are together distributed on the same concentric circle. The inner surface of the outer shell is tightly connected with the outer surface of the magnetorheological elastomer by a bond. The inner surface of the magnetorheological elastomer is stuck into a whole body with the inside iron core by the bond. The coils are divided into four groups and are respectively twisted on the internal iron core. The invention is prepared by using the characteristic that the magnetic shearing rigidity of the magnetorheological elastomer is changed with the changing of the current, and has the advantages of simple structure, good vibration absorption effect, broad applicable frequency bands(using frequency bands of several HZ to scores of HZ), strong environment adaptive capacity(resisting oil stain and dampness), good stability (the control effect is not reduced after being used for a long period), etc.

The invention discloses a changeable parallelogram crawler-type in-pipe mobile operation robot. An existing pipe robot adapts to diameter change and angle change of a pipe only through a buffer mechanism, and the adaptive capacity defect exists. The changeable parallelogram crawler-type in-pipe mobile operation robot comprises a vehicle body module, double-slide-block suspension modules, and three changeable parallelogram crawler modules. The three changeable parallelogram crawler modules tightly press the inner wall of the pipe, each changeable parallelogram crawler module is connected with the vehicle body module through one double-slide-block suspension module, and the changeable parallelogram crawler modules spread or contract relative to the vehicle body module. Each double-slide-block suspension module comprises two double-slide-block mechanisms oppositely arranged, wherein each double-slide-block mechanism comprises a first crank, a first connecting rod, a second crank, a second connecting rod and a slide sleeve. The changeable parallelogram crawler-type in-pipe mobile operation robot can automatically adjust a crawler shape and accordingly and effectively improves the in-pipe unknown-barrier-crossing capacity and passing ability in various bent pipes and reducing pipes.

The invention discloses a variable air volume room temperature control method based on fuzzy PID and a prediction control algorithm. The method comprises the steps that the temperature deviation between an expected room temperature and an actual room temperature is input to a fuzzy adaptive PID controller, PID parameter incremental quantity of the opening degree of a tail end air valve is obtained through calculation, and calibrated opening degree of the tail end air valve is output; the opening degree of the tail end air valve of the fuzzy adaptive PID controller is subtracted from the opening degree of the tail end air valve of a previous moment by utilizing a time delay link to obtain the variable quantity of the opening degree of the tail end air valve; the prediction control algorithm is adopted to obtain the actual opening degree of the tail end air valve through the variable quantity of the opening degree of the tail end air valve, and the room temperature is maintained at a set value and is kept unchanged. The control method combining the fuzzy PID with the prediction control algorithm is adopted to solve the technical problems that parameter setting difficulty is large, working condition adaptive capacity is poor, and adaptive ability is weak in an air conditioning system in the prior art.