148results about How to "Reduce search time" patented technology

The invention discloses an active power distribution network two-stage island division method and system considering flexible resource support. The method comprises the following steps: establishing a first resource allocation model considering energy storage and controllable load, the first resource allocation model comprises energy storage battery charging and discharging upper and lower limit constraints, an energy storage battery degradation cost model, a controllable load model and a controllable load related user protocol; obtaining parameter data of a target power distribution system, wherein the parameter data comprises a line parameter, a load level, a load level, a network topology connection relationship, a system operation voltage level and branch current limitation, a system fault position, a reference voltage and a reference power initial value; and according to the parameter data, establishing a mathematical model for two-stage island power supply recovery based on a first resource allocation model. The method for solving the island is simple and convenient, the island recovery amount can achieve the optimal recovery effect, meanwhile, the search time is greatly shortened, the solving efficiency is higher, and the rapid power supply recovery capacity of the power distribution network is improved.

The invention discloses an unmanned ship multi-path optimization method for large-scale dynamic search and rescue tasks, and belongs to the technical field of unmanned ship path planning. According to the method, a multi-track optimization framework combined with a deep reinforcement learning algorithm is introduced, so that the efficiency and accuracy of path planning are effectively improved. Through a multi-trajectory sampling technology, the problem of trajectory data sampling deviation in a traditional method is solved, and the robustness and adaptability of an algorithm in a dynamic rescue environment are enhanced. Meanwhile, an encoder-decoder structure is designed, path exploration of the unmanned ship is optimized by utilizing reinforcement learning, and the exploration capability of the model to a complex solution space is greatly improved. According to the method, the optimal searching path can be quickly found, and the unmanned ship is prevented from falling into a local optimal solution. According to the invention, the unmanned ship can intelligently adapt to a large-scale emergency rescue scene, the search time is greatly shortened, the task success rate is improved, and key technical support is provided for the application of the unmanned ship in emergency rescue.

The invention, which relates to the technical field of artificial intelligence, discloses a warehouse-in and warehouse-out path planning method and system for intelligent warehousing, and the method comprises the steps: generating an optimal operation path of a to-be-operated carrier through a path planning model; the method comprises the following steps: predicting whether a current carrier has a collision risk with all carriers in a warehouse in a time dimension or not through a time block, predicting whether the current carrier has a collision risk with all carriers in a warehouse in a space dimension or not through a space block, and when the time block and the space block simultaneously detect that the current carrier has the collision risk at a certain moment, judging whether the current carrier has the collision risk; the decision block judges that the current carrier has a collision conflict, and determines an optimal avoidance strategy through an action value function; according to the optimal avoidance strategy, the avoidance waiting carrier is determined, the total waiting time and the waiting space position of the avoidance waiting carrier are calculated, and the avoidance strategy is executed. According to the technical scheme, the problem that in-out garage path planning and collision processing cannot be intelligently and accurately carried out in the prior art can be solved.

The invention discloses a mobile phone artificial intelligence tour guide explanation system, which relates to the technical field of tour guide, and comprises a scenic spot search module used for searching scenic spot information throughout the country, and a user can search scenic spots needing to be visited through the module; the big data intelligent processing module automatically generates explanation content of a corresponding scenic spot by using an artificial intelligence algorithm, the voice selection module provides diversified voice types for a user to freely select and switch, the scanning explanation module has a scanning introduction character recognition function, and the voice requirement explanation module realizes voice interactive explanation service; the character input explanation module inputs content in a character form to generate personalized explanation content; the positioning module is used for acquiring the geographic position of a user in real time. According to the invention, intelligent voice explanation is realized, and different voice types can be selected according to user requirements; a user can quickly obtain detailed explanation by asking questions through voices or characters anytime and anywhere without complex operation.

The invention discloses a judicial field class case retrieval method and system based on a large model. The method comprises the following steps: constructing a judicial document vector database by combining a vector model with an obtained judicial document; performing semantic recognition and question enhancement on judicial case or judicial question texts through the large model; generating vectors for the enhanced judicial case or judicial question text through a vector model, and performing similarity retrieval in a judicial document vector database to obtain a primarily screened judicial document set; reordering the primarily screened judicial document set through a Reranker model to generate a finely arranged judicial document set; and inputting the enhanced judicial case or judicial problem text and all judicial documents in the finely arranged judicial document set into a large model for correlation analysis, outputting the judicial documents most related to the judicial case or judicial problem text, and sorting the judicial documents according to similarity to form a data set as a retrieval result. According to the method, the accuracy and efficiency of class case retrieval can be improved.

The invention provides a control method of a double-closed-loop direct-current speed regulation system based on an improved parrot algorithm, and relates to the technical field of industrial control. Comprising the steps that weights and thresholds of a BP neural network are optimized through an improved parrot algorithm, an initial optimized BP neural network is obtained, the weights for BP neural network optimization are 36, and the thresholds for BP neural network optimization are 12; according to a gradient descent principle, training the initial optimized BP neural network to obtain a final optimized BP neural network; calculating the parameters of the PID controller by using the final optimized BP neural network to obtain the optimal group parameters of the PID controller; and controlling the double-closed-loop direct-current speed regulation system by using the optimal group parameters of the PID controller. The problems that in the prior art, when PID controller parameters are adjusted on an electric dragging site, consumed time is long, difficulty is large, the workload is large, and a BP neural network falls into a local optimal solution are solved.

The invention relates to the technical field of monitoring and maintenance, in particular to an unmanned aerial vehicle automatic airport monitoring and maintenance system based on real-time data analysis. The system comprises a detection analysis unit, an analysis prediction unit, a time prediction unit, a flight adjustment unit and a planning maintenance unit. According to the invention, the planning maintenance unit plans the wind-sheltering path from the position point of the unmanned aerial vehicle to the position point where the foreign matter rolls down in the airport runway according to the acquired wind power data, the acquired wind speed data and the acquired wind direction data, and then the unmanned aerial vehicle flies from the moment t according to the flight attitude of the unmanned aerial vehicle needing to be adjusted according to the planned wind-sheltering path. When the unmanned aerial vehicle flies to a position point where a foreign matter rolls down in an airport runway, automatic maintenance is carried out, when the unmanned aerial vehicle flies under the conditions of reduced wind power and optimized flight attitudes according to a planned wind sheltering path, a strong wind area can be effectively avoided, the flight stability is ensured, and the unmanned aerial vehicle is safe and reliable. And a foreign matter position point is reached in the shortest time, so that the searching and positioning time is shortened, and the airport runway maintenance efficiency is improved.

The invention discloses an intelligent curriculum recommendation method based on implicit vector space, and relates to an artificial intelligence technology. Comprising the steps of 1, carrying out data collection and preprocessing, 2, constructing implicit vectors of users and courses, 3, calculating similarity and generating recommendation information: establishing a recommendation model, calculating the similarity or correlation between the users and the courses based on the implicit vectors by utilizing the recommendation model, carrying out sorting according to the similarity, generating a personalized course recommendation list for the users, and carrying out recommendation on the courses according to the personalized course recommendation list. And 4, optimization and feedback circulation: continuously optimizing the recommendation model through a machine learning algorithm, establishing a user feedback mechanism, and collecting satisfaction evaluation of a user on a recommendation result for improving the recommendation model.

The invention discloses a cell capacity combination model prediction method, which belongs to the technical field of battery management, and comprises the following steps: firstly, performing characteristic correlation analysis, then specifically adopting corresponding data of selected characteristics, and training the data. The method comprises the following steps: acquiring training data, processing the data by utilizing a wavelet transform algorithm, and performing prediction processing by respectively using LightGBM, improved DDPG and XGBoost so as to respectively acquire a prediction error A, a prediction error B and a prediction error C; according to the prediction error A, the prediction error B and the prediction error C, a prediction error quadratic sum reciprocal method is used for analysis to obtain the weight of a single model corresponding to each prediction error; weighting operation is carried out on all the models, so that a combined prediction model of WT + XGBoost + DDPG + LightGBM is formed; and finally, carrying out wavelet reconstruction and reverse normalization processing on the combined prediction model to obtain a prediction result. The technical problems that in the prior art, errors caused by a single prediction model are large, and the prediction model cannot be applied to production practice are solved.

The invention belongs to the technical field of signal capturing, and relates to a low-orbit satellite signal capturing method and device. The method comprises the following steps: acquiring a radio frequency signal sent by a low earth orbit satellite to obtain a zero intermediate frequency signal; obtaining orbit parameters, current time, current positioning, forecast entry and exit time, minimum Doppler frequency shift and maximum Doppler frequency shift of the low-orbit satellite; if it is judged that the current time is within the entry and exit time of the low-orbit satellite, a conjugate multiplication result is obtained according to the zero intermediate frequency signal; obtaining a discrete Fourier transform result according to the conjugate multiplication result, the minimum Doppler frequency shift in the exit-entry time and the maximum Doppler frequency shift in the exit-entry time; and judging according to a discrete Fourier transform result, if the capture is judged to be successful, calculating a Doppler frequency offset result obtained by capture, and outputting the Doppler frequency offset result, the pilot band of the zero intermediate frequency signal and the data segment of the zero intermediate frequency signal as capture results. According to the invention, the calculation amount of low-orbit satellite signal acquisition is small and the power consumption is low.

The invention provides a code searching method and device, electronic equipment and a storage medium, and relates to the technical field of retrieval. The method comprises the steps that in response to a received retrieval statement of a natural language, a natural language annotation corresponding to the retrieval statement is searched in a natural language annotation set of a code library, and the natural language annotation is used for indicating functions of functions in function code snippets in the code library; and if the natural language annotation is found, determining a function code snippet corresponding to the natural language annotation in the code library. According to the embodiment, codes do not need to be adopted as search terms, natural languages can be adopted for searching, searching personnel are more friendly, the searching habits of the personnel are better met, searching can be conducted through the functions, and the needed functions can be searched easily.

The invention relates to the technical field of satellite communication, and discloses a communication method and system for a long-voyage ship based on L frequency band communication, and the method comprises the steps: recognizing the navigation environment and communication demands of the long-voyage ship, recognizing the satellite signals of the long-voyage ship and a communication end through recognizing the frequency band characteristics of a communication frequency band, and carrying out the communication of the long-voyage ship; a signal tracking network of satellite signals is set, signal parameters of the satellite signals are monitored in real time according to the signal tracking network, a quality feedback mechanism of the signal parameters is set, the current signal quality grade is identified, and the signal quality requirements of the long-voyage ship and the communication end are identified based on the current signal quality grade; the method comprises the following steps: setting a multi-beam switching mechanism pointing to beams and a signal locking network of satellite signals, setting a confidential path for transmitting data, and generating a communication result of a long-voyage ship and a communication end in combination with the multi-beam switching mechanism, the signal locking network and the confidential path; according to the invention, the stability and safety of signal transmission can be enhanced, and the communication quality of the long-voyage ship is improved.

The invention relates to a Bayesian optimization-based lithium ion battery health state prediction method and system, and the method comprises the steps: 1, obtaining the charging and discharging data of a battery, carrying out the preprocessing of the charging and discharging data of the battery, and extracting a key health factor from the preprocessed data; the key health factors comprise constant-current charging time, voltage fluctuation and the change rate of discharge capacity along with time; 2, constructing an initial BiSLTM model based on the key health factors, setting an initial hyper-parameter combination, adjusting hyper-parameters by using Bayesian optimization to obtain an optimal hyper-parameter combination, and constructing a BiLSTM optimal model by using the obtained optimal hyper-parameter combination; and step 3, predicting the health state of the battery by using the BiLSTM optimal model. According to the method, the hyper-parameters of the BiLSTM model are optimized through Bayesian optimization, the problem of BiLSTM hyper-parameter tuning is solved, the globally optimal solution can be efficiently found under the condition that the number of evaluation times is limited, the search duration is greatly shortened, and the generalization efficiency of the model is improved.

The invention discloses a mobile terminal path planning method and system based on a distance field, and the method comprises the steps: firstly building an ESDF map, and calculating a gradient value of each grid point in the ESDF map; establishing a state vector and a discretization state transition equation of the mobile terminal based on the vehicle model; and finally, calculating the next state of the mobile terminal moving from the current node by using an improved # imgabs0 # algorithm until an optimal path from the starting node to the terminal node is formed. Wherein the ESDF cost is calculated by using the ESDF value and the gradient value of the node, and the cost function of the improved # imgabs1 # algorithm is the sum of the actual cost, the heuristic estimation cost and the ESDF cost. According to the distance field and gradient information of the ESDF map, collision detection and expansion cost calculation in the # imgabs2 algorithm are combined, and the path planning efficiency is improved while the path safety is guaranteed.

The invention discloses a parking lot planning method based on real-time operation big data of a parking lot, and the method can dynamically optimize the use and management of the parking lot based on the real-time operation big data of the parking lot. Firstly, in combination with vehicle access time and direction data and operator mobile phone signaling data, a trip starting point and a trip ending point of regional traffic flow are accurately calculated, and accurate analysis of traffic flow modes and parking requirements is facilitated. And secondly, a space-time state model of the parking lot is constructed through real-time flow data, the utilization rate and the turnover rate of the parking spaces in different time periods can be reflected in real time, and the parking demand and the vehicle flow trend in the future are predicted, so that a scientific basis is provided for resource optimization configuration of the parking lot. And finally, performing path planning by using the real-time vehicle position, the destination and the predicted parking lot state data, dynamically adjusting the vehicle driving route, and optimizing the optimal path for the vehicle to arrive at the destination.

The invention discloses a multi-objective optimization method for a radiotherapy plan, and the method comprises the steps: S1, building a multi-objective optimization model, and enabling the multi-objective optimization model to be used for the multi-objective intensity optimization of a radiotherapy reverse intensity modulation plan, and generating a Pareto solution; step S2, inputting a Pareto solution currently generated by the multi-objective optimization model into a weight search network model, wherein the weight search network model predicts a sub-objective weight of a next round of multi-objective intensity optimization of the multi-objective optimization model according to the current Pareto solution; s3, optimizing a single-target strength optimization problem with sub-target weights through a multi-target optimization model to update a current Pareto solution; and S4, circularly executing the steps S2 to S3 until a Pareto solution generated by the multi-objective optimization model forms a Hausdorff distance dHaus (Zin, Zout) between an upper bound and a lower bound of the Pareto in space, the distance dHaus (Zin, Zout) is less than threshold, and the threshold is a predefined threshold parameter which is greater than 0. According to the invention, a deep learning method is creatively used to solve the problem of constraint weight search in a radiation therapy reverse plan multi-objective optimization method.

The invention discloses an intelligent real estate mobile terminal drop house system and method, and relates to the technical field of real estate registration, and the system comprises a data collection module which is used for collecting preference information of a user; the intelligent recommendation module is used for recommending related real estate information according to the preference information of the user acquired by the data acquisition module; the evaluation module is used for surveying and mapping the real estate according to the information of the data acquisition module and evaluating the value and potential risk of the real estate; information of real estate can be recommended in combination with requirements and characteristics of users, and personalized recommendation can improve user satisfaction and application stickiness, increase possibility of real estate transaction, optimize user experience and reduce search time; accurate real estate data can also be provided at the mobile terminal, the user is helped to make a more intelligent decision, and the user is assisted to quickly understand various information of logging in real estate at the mobile terminal.

The invention relates to the technical field of information technology, in particular to a decentration-based peer-to-peer group social network system and a building method thereof, and the method comprises the following steps: setting network nodes, each network node having local data collection, primary processing and decision-making capabilities, the network nodes have self-organization and dynamic participation capabilities, the system adopts social communication protocol design, realizes data transmission among the network nodes through a wireless communication technology, supports flexible network node arrangement and communication coverage, adopts distributed data processing and storage, and is high in reliability and high in reliability. And the network nodes share data and carry out cooperative processing. Through autonomous interaction and data sharing among network nodes, the system can realize efficient resource utilization, real-time environment perception and dynamic decision adjustment, improve the resource allocation efficiency, realize real-time data acquisition and optimize dynamic environment response, and significantly improve the management efficiency and the user experience.

The invention discloses an air ticket calendar mode change search method and device and electronic equipment, and relates to the technical field of air ticket search or other related fields, the method comprises the following steps: responding to a change search request, splitting a target date range to obtain a date units, and on the basis of the date units, obtaining a data unit; b new itinerary flight combinations corresponding to each date unit are obtained, historical freight rate information of the original flights is obtained, the historical freight rate information comprises historical freight rates of the original flights and corresponding change rules and procedure fees, and multiple new itinerary freight rates of each new itinerary flight combination are obtained based on the change rules and the procedure fees; and calculating the change price difference of each new itinerary freight rate, obtaining the lowest price difference in each date unit based on all the change price differences, and outputting a calendar mode change search result containing flight information corresponding to all the lowest price differences. According to the invention, the technical problem that the user can only obtain the air ticket change search result of the fixed date in the prior art is solved.

The invention provides a frequency point selection method, which is used for selecting a frequency point for a terminal in a communication system so as to determine a working frequency band to receive a broadcast signal of a satellite or a base station, and comprises the following steps in each frequency point selection process: acquiring frequency point information, the frequency point information comprises a plurality of standard frequency points and historical frequency points used last time by the terminal; searching a frequency correction channel of the historical frequency point and calculating signal intensity, and if the signal intensity of the frequency correction channel of the historical frequency point is greater than or equal to a preset threshold, using the historical frequency point; and otherwise, searching in a plurality of standard frequency points except the historical frequency points one by one until the signal strength of the frequency correction channel of one standard frequency point is greater than or equal to the preset threshold value, and using the standard frequency point. According to the technical scheme provided by the invention, the invalid frequency point searching and selecting process is avoided by preferentially searching the frequency correction channel of the historical frequency points, so that the frequency point selecting time is shortened.

The invention discloses an automobile central control intelligent driving assistance method and system, relates to the field of intelligent driving assistance, and is used for solving the problem that a driver is difficult to quickly find a required function in a central control display screen in a specific scene, obtaining each functional block, carrying out current feature state adjustment scoring on each functional block and carrying out statistical calculation, and finally obtaining a final functional block. The sorting important features comprise environment feature information and driver feature information, and the sorting important feature of each functional block is obtained by combining multi-source data and Kalman filtering; the to-be-adjusted function block and the historical click preference degree combination in the corresponding environment of the user are subjected to matching scoring, a new adjustment score of each function block is obtained, the new adjustment score of each function block is compared with a preset function block sorting threshold value, the adjustment position of each function block is obtained, the operation convenience of the central control display screen is improved, and the user experience is improved. And the searching time of the user on the central control screen is shortened.

The invention relates to the field of semiconductor equipment, in particular to a set parameter optimization method, computing equipment, a medium and a semiconductor machine system, and aims to solve the problem of low efficiency of an existing point-by-point sampling method. In order to achieve the purpose, the set parameter optimization method is applied to a semiconductor machine, the semiconductor machine has delayed feedback after set parameters are adjusted, and the method comprises the steps that the optimization range of the set parameters is obtained; the set parameters are adjusted in the optimization range, and feedback signals are collected; and determining the optimal value of the set parameter in the optimization range based on the change trend of the feedback signal. According to the method, parameter optimization is set based on the change trend of the feedback signal, point-by-point sampling is not needed, the sampling frequency is reduced, the optimization time is shortened, and the production efficiency is improved.

The invention discloses a satellite network access method and device, electronic equipment and a storage medium, and relates to the technical field of mobile communication. The method comprises the steps that a network access request of terminal equipment is generated through a ground network, the ground network is a mobile network deployed on the ground, and the network access request represents that the terminal equipment requests to be accessed to a satellite network; detecting whether the terminal equipment meets an access condition of a satellite network through a ground network; and under the condition that the terminal equipment meets the access condition of the satellite network, controlling the terminal equipment to access the satellite network according to a negotiation result of the ground network and the satellite network, or controlling the terminal equipment to search the satellite network according to a search range returned by the satellite network, and accessing the satellite network after the satellite network is searched. The technical problems that in the prior art, when terminal equipment of a user accesses a satellite network from a ground network, network search time is long, and satellite network access efficiency is low are solved.

The invention provides a message processing method and device, electronic equipment and a storage medium. The method is applied to network equipment, and comprises the following steps: after an MPLS message carrying an MPLS label stack comprising multiple layers of MPLS labels is received, generating a label bit map according to the multiple layers of MPLS labels and a preset generation rule; for a target MPLS label corresponding to a bit with a value of 1 in the label bit map, searching a preset MPLS label forwarding table in parallel to obtain a search result corresponding to the target MPLS label; and if a target search result indicating the next hop forwarding operation exists in the search results corresponding to the target MPLS label and the target search result is not the first search result, processing the MPLS message according to the search result in front of the target search result and the target search result in sequence. According to the invention, the processing performance of the MPLS message can be improved.

The utility model provides a floating type monitor for detecting pollution gas, which relates to the technical field of gas monitoring equipment and comprises a floating base, an arc-surface base station is fixedly mounted at the top of the floating base, an assembly sleeve is fixedly mounted in the floating base, a distance measuring window is fixedly connected in the assembly sleeve, and the distance measuring window is fixedly connected with a gas outlet of the floating base. A sealing ring is embedded between the inner surface wall of the assembling sleeve and the outer surface wall of the distance measuring window, an inner open groove is formed in the floating base, a partition plate is fixedly installed in the inner open groove, and a set of driving motors are fixedly installed in the floating base. The data obtained by the detection element is used as an adjustment basis, the equipment is controlled to continuously move towards the shallow water area, and in cooperation with the built-in positioning element, when the equipment is recycled, related personnel only need to reach the target shallow water area, so that the equipment searching time can be greatly shortened, and the equipment recycling cost is reduced.

The invention belongs to the field of computer science and control, and particularly relates to a region full-coverage path searching method which comprises the following steps: S1, defining a grid information coding rule; s2, defining a grid division rule of the map; s3, defining a grid merging rule; s4, according to the rules in the step S1, the step S2 and the step S3, performing grid division and merging on the map; s5, searching an optimal path of the 2 * 2 grids by using a depth-first search method, and combining the optimal path with other 2 * 2 grid paths of the current layer; and S6, according to the coverage path of the previous layer, carrying out starting point and end point division on each group of 2 * 2 grids of the current layer, and repeating the step S5. According to the method, the search time can be remarkably shortened, the reliability and the stability of the planned path are improved while full coverage is ensured, and the problem that the DFS algorithm cannot be solved in time can be effectively solved.

The invention provides an unmanned aerial vehicle routing inspection method and system based on semantic guidance and an asset value cost graph, and the method comprises the steps: reasoning a possible distribution region of a target when the target is not directly observed through analyzing the structure law of semantic information, such as an equipment number, a function partition identifier and the like, in a transformer substation; generating a confidence map to guide the unmanned aerial vehicle to efficiently inspect; constructing a semantic cost graph fusing the static asset value and the dynamic operation risk, and quantifying potential loss of different regions when collision or falling occurs; according to the real-time health state of the unmanned aerial vehicle, semantic guidance is emphasized during normal inspection through a dual-guidance function fusion and switching mechanism, and the unmanned aerial vehicle is automatically switched to asset value cost minimization guidance in an abnormal or emergency state to realize safe disposal; the system adopts a framework of semantic reasoning, cost evaluation and flight control module decoupling, and guidance intelligence and control reliability are ensured. According to the invention, the inspection blindness of the unmanned aerial vehicle is obviously reduced, and the abnormality handling risk is reduced.

The invention discloses an intelligent auxiliary vehicle searching method and system based on a mobile robot. The method comprises the following steps: acquiring attribute information and position information of a to-be-searched vehicle; vehicle searching path planning is carried out based on an improved A * algorithm, and an initial vehicle searching path is obtained; carrying out smoothing processing on the initial vehicle searching path; based on a dynamic obstacle avoidance TEB algorithm, designing a reward or penalty function for each influence factor by configuring the weight of each influence factor of an obstacle, a speed and an acceleration, scoring feasible routes, and selecting a local route with the highest score to send a speed instruction; the acceleration and the speed of the robot are restrained, so that the robot bypasses the obstacle to reach the target point fastest under kinematics constraint; the mobile robot can be used for assisting the user in querying the vehicle and autonomously guiding the user to complete rapid vehicle searching, compared with an intelligent vehicle searching platform or applet put in the current market, the experience feeling is better, the vehicle searching efficiency is higher, and equipment safety and use specifications are guaranteed.

The present disclosure provides a cleaning apparatus for a surface submerged in a liquid, including: an apparatus body; and an automatic rising-sinking structure, where through an operation of the automatic rising-sinking structure, the apparatus body rises or sinks. The present disclosure has the following beneficial effects: The cleaning apparatus for the surface submerged in the liquid automatically rises and sinks through the automatic rising-sinking structure, without being manually searched and lifted out of a water surface. The cleaning apparatus for the surface submerged in the liquid can also automatically rises out of the water surface when a filter box is to be cleaned or machine maintenance is required. The present disclosure reduces a manual labor, shortens searching time, and improves a working efficiency.

The present invention relates to the field of sonic insecticide, and specifically to a sonic insecticide device in a space capsule, wherein the sonic insecticide device is composed of a plurality of sonic insecticide units and a connecting device, and can be spliced ​​and assembled according to the size of the area actually requiring insecticide elimination. The sonic insecticide device surrounds the insect elimination area, and the ultrasonic waves generated by the sonic insecticide unit can be transmitted in the connecting device and propagated outward through a plurality of through holes. The sonic insecticide device in the space capsule, through the combined use of the sonic insecticide unit and the connecting device, realizes three-dimensional insect elimination in the insect elimination area, expands the area affected by the sonic wave, and enhances the insect elimination effect.