54 results about "Semiconductor factory" patented technology

The invention discloses a semiconductor wafer manufacturing factory logistics jam prediction method and system, and relates to the technical field of semiconductor manufacturing process optimization, and the method comprises the steps: collecting and preprocessing historical logistics data of a semiconductor manufacturing intelligent factory, carrying out the spatial-temporal feature fusion of the preprocessed data, generating high-dimensional representation data, and carrying out the prediction of the logistics jam of the semiconductor wafer manufacturing factory. Performing coding feature extraction processing on the high-dimensional representation data, extracting spatio-temporal features through a time sequence dynamic sensing operation and a spatial relationship self-adaptive operation, performing spatio-temporal feature fusion processing, performing cross-sequence feature integration on a processing result and complex spatio-temporal features, and performing decoding prediction processing on the integrated features, and through a causal constraint attention mechanism and cross-sequence correlation modeling, predicting a logistics state matrix sequence, and based on the logistics state matrix sequence, carrying out congestion state determination. According to the method, high-precision prediction and quick response of logistics blockage of the semiconductor factory are realized, and the problems of spatial-temporal characteristic splitting, response lag and high false alarm rate of a traditional method are solved.

The invention discloses a three-dimensional modeling device and method for a semiconductor factory. The method comprises the steps that an acquisition module acquires interface layout information and connection element data of the semiconductor factory; the equipment modeling module receives semiconductor factory interface layout information, determines a three-dimensional modeling starting point and generates a semiconductor factory equipment model containing an equipment interface; the connection element modeling module determines an element connection path by using semiconductor factory starting equipment or factory service interfaces, combining element interface matching rules and logic relation information among elements, and fusing a three-dimensional touch automatic checking algorithm and an optimal path algorithm based on safe distance correction; and interface connection among elements in each semiconductor factory equipment model in the equipment modeling module is realized, and three-dimensional modeling of the semiconductor factory is completed. According to the method, the layout information and the connection element information are fused, and three-dimensional modeling of the semiconductor factory is efficiently and accurately achieved through the logic relation between the connection elements and a three-dimensional touch automatic verification algorithm.

This invention discloses a device and method for generating a simulation system for pure water preparation in a semiconductor factory. The device includes: a process generation module, which acquires raw water parameters and pure water requirement parameters, and generates a pure water preparation process path and process parameters based on a pure water process selection unit; and a dynamic pure water preparation simulation system built based on the pure water preparation process path and process parameters; and a parameter configuration module, which embeds a functional material configuration quantity calculation unit into the dynamic pure water preparation simulation system, and configures the functional materials for each process by combining process parameters, pure water requirement parameters, and basic properties of functional materials. The generation device provided by this invention can significantly improve the efficiency and parameter accuracy of pure water preparation process design in semiconductor factories, calculate the optimal configuration quantity of functional materials required for each process, and reduce the construction cost of the pure water preparation system in a semiconductor factory.

The invention relates to the technical field of industrial automation and semiconductor manufacturing, in particular to a semiconductor factory material distribution control system and method based on multi-system collaboration, which can realize full-process automatic closed-loop control of material distribution by integrating a WMS, an MES, an OA system and an AGV / PDA device. Comprising the steps that after an MES work order module generates a work order, a WMS work order module is triggered to distribute material batches, an AGV control module dynamically dispatches and distributes the work order to a workshop, a PDA device scans a code, checks and synchronizes data, and a related department is notified through an OA cooperation module when a task is completed or abnormal, so that the problems of information island, much manual intervention, response lag and the like in a semiconductor factory are solved. And the material distribution efficiency and the cross-department cooperation capability are improved.

The present application relates to a semiconductor factory production data analysis and presentation method and system based on multi-agent cooperation. First, access to multi-source heterogeneous production data and standardization processing. Subsequently, a plurality of serial special ability agents are executed in turn: data agent classifies and stores data; aggregation analysis agent performs cross-system calculation to generate comprehensive indicators; scene recognition agent matches real-time data with preset rules to identify the current running scene; key parameter recommendation agent outputs a sorted key parameter list according to the scene. Finally, the scene, indicators and parameter list are integrated and visualized through a leadership cockpit with natural language interaction function, and can respond to user queries to trigger reanalysis. Through agent cooperation, the present application realizes automatic data processing, scene recognition and dynamic key parameter recommendation, and solves the technical problem of quickly focusing on core production problems from isolated data.

The invention discloses a crown block track surface cleaning device in a semiconductor workshop, which comprises a bottom plate, a frame is arranged on the bottom plate, and a first negative pressure fan and a second negative pressure fan are arranged in the frame; a negative-pressure air draft longitudinal cylinder extending upwards is arranged in the frame, and the bottom end of the negative-pressure air draft longitudinal cylinder is connected with a first negative-pressure fan and a second negative-pressure fan through pipelines in a communicating mode. A negative pressure air draft transverse barrel is arranged at the top end of the negative pressure air draft longitudinal barrel, the negative pressure air draft transverse barrel is vertically arranged at the top end of the negative pressure air draft longitudinal barrel, and the negative pressure air draft transverse barrel is communicated and connected with the negative pressure air draft longitudinal barrel; the two ends of the negative-pressure air draft transverse barrel are provided with dust removal modules which protrude downwards and are matched and symmetrical, and each dust removal module comprises a dust removal head cavity. Through non-contact suspended positive and negative air passage cleaning, efficient and automatic cleaning of the crown block rail surface of the semiconductor workshop is achieved, and the cleanliness of the semiconductor production environment is effectively improved.

This invention discloses a method for predicting energy data in semiconductor factories based on the fusion of multiple time-series models. The method includes: cleaning and defining the features of energy data; selecting and training multiple individual models based on the data characteristics of the processed energy data; selecting a suitable individual model based on scenario discrimination indicators, wherein the scenario discrimination indicators include fluctuation amplitude and seasonal cycle; the fluctuation amplitude is calculated by the standard deviation of the data or the range within a moving window, and the seasonal cycle is obtained by detecting the cycle strength using an autocorrelation function; fusing the prediction results of the selected models; evaluating the fused model and making energy predictions based on the evaluation results. This invention can adaptively select the optimal model, ensuring the accuracy and stability of energy prediction.

A system for monitoring one or more conditions of an automation system of a semiconductor factory includes one or more instrumented substrates, one or more sealable containers and one or more system servers. The one or more instrumented substrates include one or more sensors. The one or more sensors measure one or more conditions of the one or more instrumented substrates as the one or more sealable containers transport the one or more instrumented substrates through the semiconductor factory. The one or more sealable containers also receive sensor data from the one or more sensors included on the one or more instrumented substrates. The one or more system servers are configured to receive the sensor data from the one or more sealable containers. The one or more servers are configured to identify one or more deviations in the measured one or more conditions.

The invention relates to the technical field of electronic water pumps, in particular to a liquid-cooled electronic water pump which comprises a pump base, a pump body arranged at the top end of the pump base and a pump head installed at one end of the pump body, the electronic water pump can conduct liquid cooling heat dissipation so as to be free of external environment limitation, liquid cooling is higher than traditional air cooling efficiency, the temperature of a motor can be accurately controlled, and the service life of the motor is prolonged. Operation heat is absorbed by liquid in the sealing cover and then is discharged through the leading-out pipe, the environment temperature and humidity are not interfered, the electronic water pump is matched with scenes such as semiconductor plants and data machine rooms, and meanwhile through the arrangement of the waste heat dissipation assembly, after the electronic water pump is stopped, cooling liquid in the water drainage box is sprayed to a pump body and cooling fins through a water drainage head of a liquid drainage pipe for waste heat dissipation; the cooling liquid after heat dissipation finally flows into the receiving box to be collected, the whole process is uninterrupted, the temperature of the motor in the pump body can be slowly reduced to a safety interval, damage of waste heat to core components of the pump body is effectively avoided, the overall service life is prolonged, and the operation stability during follow-up starting is guaranteed.

This invention relates to a micro-factory equipment for semiconductors, comprising: an atmospheric pressure processing device; an atmospheric pressure equipment housing, wherein the atmospheric pressure processing device is disposed within the atmospheric pressure equipment housing; the atmospheric pressure equipment housing includes an inner protective layer mechanism and an outer protective layer mechanism, both of which are heat-insulating protective layer structures with metal layers, and an airflow buffer layer is formed between them; a temperature compensation unit connected to the processing area where the atmospheric pressure processing device is located, used to compensate for temperature fluctuations caused by changes in equipment power within the processing area; and a dust filtration unit disposed within the atmospheric pressure equipment housing. Through the above configuration, the problem of huge investment in the construction of cleanroom and constant-temperature workshops in current semiconductor factories can be solved.

The invention discloses a capacity management and control device and method for a semiconductor factory service system, and the device comprises a demand end module which is used for obtaining the position information and consumption demand information of machine equipment in a semiconductor factory; the supply end module is used for acquiring pipeline position information and pipeline supply capacity of a factory affair system in the semiconductor factory; the capacity management and control module is used for receiving the machine equipment position information given by the demand end module and the pipeline position information given by the supply end module, realizing column network matching of each machine equipment and each pipeline, and receiving the use amount demand information given by the demand end module and the pipeline supply capacity given by the supply end module, and the capacity management and control of the semiconductor factory service system is realized in combination with a point location distribution mode. According to the method, the power capacity of the whole factory is comprehensively and dynamically controlled through data communication between the factory affair system in the semiconductor factory and the machine equipment in the spatial dimension, and cost waste caused by mismatching of the demand end and the supply end is avoided.

The invention provides a production schedule Gantt chart drawing method, device and equipment for a semiconductor factory, and belongs to the field of semiconductor manufacturing, and the method comprises the steps: receiving work order schedules of all processing machines, and determining the time granularity of a time axis of a Gantt chart according to the work order schedules; constructing a data structure hierarchy corresponding to the processing machine to display the association between the business scenes; receiving a real-time Gantt chart fed back by rendering, receiving interface data corresponding to each processing machine, calculating the interface data, and drawing task bars forming a business scene according to time granularity through each Gantt chart component to obtain a real-time Gantt chart corresponding to the work order schedule; and displaying the real-time Gantt chart so as to carry out dragging modification on the task bar, and outputting the real-time Gantt chart after dragging verification. According to the processing scheme provided by the invention, different Gantt chart levels are designed for different types of machine processing flows in a semi-conductor factory, and an association relationship is established; and the page time axis granularity can be adjusted, and more comprehensive information can be displayed.

The application discloses an Ethernet and serial communication conversion device and method for semiconductor equipment, which comprises a hardware system and a software system, the hardware system contains an industrial-grade main controller, a plurality of independent configurable serial communication interfaces, an Ethernet interface, a state indication and alarm module and the like; the software system runs on the main controller and comprises a bidirectional protocol conversion module, a data priority processing and flow control module and a state monitoring and fault diagnosis module. The data priority processing and flow control module is internally provided with a data processing mechanism based on static priority, or adopts an adaptive priority dynamic re-marking and flow shaping method based on process stage sensing and queue pressure prediction, and the priority of a data frame and a token bucket injection rate are dynamically adjusted. The application can adapt to the harsh environment of a semiconductor factory, realize high-reliability and low-delay bidirectional conversion of a multi-protocol serial device and an Ethernet, and has perfect state monitoring and fault automatic recovery functions.

The application relates to a safety warning method, device and equipment for a semiconductor factory work scene. The method comprises the following steps: acquiring a work scene image of a semiconductor factory; inputting the work scene image into a target detection model and receiving a detection result of the target detection model; the detection result comprises a first detection result corresponding to a first target detection head and a second detection result corresponding to a second target detection head; determining an isolation index according to the first detection result and the second detection result; and determining safety warning information of the work scene according to the isolation index. The method can improve the warning accuracy.

The present disclosure provides a method for process control for a semiconductor plant, a higher-level computer, a controller and a system, and relates to the field of semiconductor technology.The process control procedure for a semiconductor plant comprises: capturing a process file required to control the semiconductor plant to execute a predefined process sequence; the process file being generated based on received process configuration parameters and comprising several process commands corresponding to all process operations required to execute the predefined process sequence, with each process command being assigned to a process operation; transmitting the process file to the controller at the start of the process so that the controller sequentially analyzes the process commands individually and controls the semiconductor plant based on the process parameters in the process commands to execute the corresponding process operation.The present disclosure enables controller-centric process control, increases the reliability of process control, improves the efficiency of information transfer, reduces plant costs and increases the stability of process control.

The invention discloses an electrical load prediction method, system and device for a semiconductor factory and a medium. The method comprises the following steps: acquiring historical electrical load data of a power grid system of the semiconductor factory; determining a lagging order according to the historical electrical load data, and screening input features according to the lagging order; a tree model algorithm is adopted to fit the input features and the electricity consumption in a period of time in the future, and a prediction model is constructed; and predicting future electric load data of the power grid system through the prediction model and outputting a result. According to the method, feature screening is carried out through hysteresis correlation, secondary screening is carried out on the features through feature importance of the tree model, and the autoregression prediction accuracy is improved.

The utility model discloses a semiconductor factory affair comprehensive waste gas treatment device which comprises an oxygen catalyst of an existing structure, at least one waste gas treatment mechanism connected with the oxygen catalyst and an ionization treatment mechanism which are connected in sequence, the waste gas treatment mechanism comprises a shell, and a containing cavity is formed in the shell. An inlet is formed in the first side of the lower bottom face of the shell, and an outlet is formed in the second side of the top of the shell. An air guide pipe is integrally formed at the inlet towards the interior of the containing cavity, a plurality of air holes are formed in the end of the air guide pipe, solid filler or liquid filler is arranged in the containing cavity, and a bent part is arranged on the air guide pipe so that the air holes can be located in the solid filler or the liquid filler. The waste gas treatment mechanism adopted by the utility model can react with solid or liquid fillers in a multi-stage manner, so that the waste gas treatment efficiency is improved, and the device is suitable for organic and inorganic comprehensive waste gas.

The application relates to a safety warning method, device and equipment for a semiconductor factory work scene. The method comprises the following steps: acquiring a work scene image of a semiconductor factory; inputting the work scene image into a target detection model and receiving a detection result of the target detection model; the detection result comprises a first detection result corresponding to a first target detection head and a second detection result corresponding to a second target detection head; determining an isolation index according to the first detection result and the second detection result; and determining safety warning information of the work scene according to the isolation index. The method can improve the warning accuracy.

The invention discloses a semiconductor factory fault tracing method and system based on LLM and retrieval enhancement, belongs to the field of semiconductor manufacturing and intelligent troubleshooting, and aims to solve the problems that historical processes and defect report fields of a semiconductor factory are not uniform, a retrieval mode is single, the accuracy rate is low, and a fault tracing result lacks causal reasoning and interpretation. Firstly, report key information fields are extracted through combination of a large language model, rules and codes, and source identifiers are added; secondly, field semantic normalization is achieved through large language model cue word mapping, and field data structure specifications are unified; then writing the data into an elastic search database, defining text type fields and adding comprehensive content fields containing field combination information; when a user asks a question, the question is analyzed through a large language model to generate a selected statement for query, and two-stage retrieval is executed. According to the method, the retrieval accuracy and recall rate can be improved, the interpretability of fault tracing is realized, and the method has dynamic expansion capability and assists engineers in quickly removing faults and making decisions.

The application discloses a kind of semiconductor factory IPA-containing wastewater recycling device, it is related to semiconductor industry wastewater resource recovery technical field, including premixing tank, heat exchanger, rectifying column, buffer tank, osmotic vaporization membrane assembly, crude IPA tank, resin refining tank and finished product IPA storage tank.Premixing tank is provided with flow guide pipe, paddle assembly and rotating shaft stem, can drive wastewater circulation flow, realize raw material homogeneous premixing;Rectifying column bottom is provided with material scraping assembly, and material scraping assembly includes scraping shaft, scraping rod and spiral feeding auger, can be scraped to the deposition kettle residue at tower bottom and is transported to waste storage tank inside centralized collection and discharge;Premixing tank and material scraping assembly are linked by transmission mechanism, realize power sharing.The application can improve wastewater premixing effect and rectifying operation stability, reduce kettle residue deposition and shutdown cleaning frequency, pre-adjust pH value, realize the efficient resource recycling of semiconductor industry IPA-containing wastewater.

The invention discloses a device and a method for treating semiconductor fluorine-containing wastewater and domestic sewage. The device comprises a primary sedimentation tank, a secondary sedimentation tank, a middle water tank, a domestic sewage collecting device, a mud-membrane symbiotic SBR (sequencing batch reactor), a first peristaltic pump, a second peristaltic pump, a third peristaltic pump, a fourth peristaltic pump, a first stirrer, a PLC (programmable logic controller) integrated control center, a suction device, a third stirrer and a second stirrer. After being subjected to calcium salt precipitation two-stage treatment in the primary precipitation tank and the secondary precipitation tank, the fluorine-containing wastewater enters the intermediate water tank through a second peristaltic pump; starting the first stirrer, starting the suction device at the same time, and discharging the precipitate to the collection device for independent treatment; in the short-cut nitrification starting period, actual domestic sewage of a semiconductor factory is used as inlet water. The method has the advantages that NOB is difficult to enrich again under the inhibition of fluorine ions, short-cut nitrification is high in starting speed and more stable, the fluorine content of effluent is far lower than the national emission standard after primary treatment and biological treatment, and the treatment cost is low.

The invention discloses a semiconductor factory power demand simulation device and method considering scheduling and supply, and the device comprises a production scheduling module which is used for obtaining semiconductor production task information, and giving a processing sequence of each semiconductor production device; the equipment energy consumption module is used for acquiring the processing sequence, given by the production scheduling module, of each semiconductor production equipment, simulating the processing behavior of each semiconductor production equipment, and giving sequential power demand data of each semiconductor production equipment; and the pipe network supply module is used for acquiring the time sequence power demand data of each semiconductor production device given by the device energy consumption module, simulating a pipe network operation behavior according to the pipe network layout, and giving time sequence pipe network operation state data. By constructing a semiconductor factory power demand simulation model of three levels of production scheduling-equipment energy consumption-pipe network supply, accurate prediction of the power demand of the semiconductor factory is realized, and the problem that power demand prediction and planning of the semiconductor factory are disjointed is solved.

1. The name of the design product: dual-vision thermal imaging monitor (CTD-01). 2. The use of the design product: dual-spectrum online monitoring equipment for online monitoring and safety warning of electrical cabinets in semiconductor factories. 3. The design points of the design product: the combination of shape and pattern. 4. The picture or photo that best indicates the design points: perspective view.

The application discloses a kind of semiconductor secondary configuration-oriented multi-source document intelligent question and answer method and device, method includes: obtaining user input information;By matching analysis to user input information and preset scene library, determine question and answer scene;Parse user input information, extract each target to be indexed, obtain index demand;Combining semiconductor factory hierarchical knowledge graph and index demand, determine multiple index objects;Based on index demand, and by multiple index mode in each index object is indexed, give multiple index results;Fusion question and answer scene and multiple index results, obtain the first question and answer result of each proxy point;The first question and answer result of each proxy point is context enhanced, obtain target question and answer result.Semiconductor factory hierarchical knowledge graph is constructed by the data characteristics of semiconductor secondary configuration, fusion question and answer scene and index demand, realize fast cross-document association query and traceable information retrieval, improve answer quality.

The application discloses a kind of semiconductor factory power demand simulation device and method considering scheduling and supply, device includes production scheduling module for obtaining semiconductor production task information, give the processing sequence of each semiconductor production equipment;Equipment energy module is used to obtain the processing sequence of each semiconductor production equipment given by production scheduling module, simulate the processing behavior of each semiconductor production equipment, give the time-sequenced power demand data of each semiconductor production equipment;Pipe network supply module is used to obtain the time-sequenced power demand data of each semiconductor production equipment given by equipment energy module, and according to pipe network layout, simulate pipe network operation behavior, give time-sequenced pipe network operation state data.Through the construction of "production scheduling-equipment energy-pipe network supply" three levels of semiconductor factory power demand simulation model, accurate prediction of semiconductor factory power demand is realized, and the problem that semiconductor factory is disconnected for power demand prediction and plan is solved.

The application relates to the technical field of semiconductor factory environment monitoring, and discloses a mobile environment parameter monitoring system and method based on a semiconductor chip factory, which comprises the following steps: first, dividing a monitoring coverage area into a plurality of preset areas connected by a mobile monitoring path; when a monitoring trolley enters a first preset area along the path, a PDA device on the trolley communicates with corresponding area monitoring points to obtain a measurement task; after a measurement period and a measurement frequency are configured, the environment parameters are measured to obtain results; then, the environment of the remaining preset areas is measured in sequence along the path; finally, the measurement results of the areas are summarized, and the monitoring results of the environment parameters of the whole factory area are output. The application realizes linkage between the monitoring trolley and the division of the factory area, standardizes the measurement standard, coordinates the measurement progress of multiple areas, improves the monitoring and data summarization efficiency, and meets the actual needs of the environment monitoring of the whole factory area of the semiconductor factory.

The utility model relates to a fork truck device for mobile special gas equipment cabinet, including first main body frame part, second main body frame part, telescopic connecting part, lifting tray part and control part. Its advantage lies in, and the first main body frame part and second main body frame part realize transverse extension through telescopic connecting part, and this design makes the device can be compatible with different specifications of special gas cabinet, especially suitable for the transfer demand of multiple size special gas cabinet in semiconductor factory, improves the application range of overall device, in addition, through setting tray lifting part on the first main body frame part and second main body frame part, can realize the automatic lifting to tray, compared with the lifting mode of small -size manual fork truck swing hydraulic arm, the device automatic lifting tray, improves the automation degree of lifting device, reduces the labor intensity of staff.

The invention provides a semiconductor factory point location data prediction method and device based on correlation calculation, and belongs to the field of semiconductor intelligent manufacturing, and the method specifically comprises the steps: obtaining historical sampling data corresponding to each physical point location in a data granularity and semiconductor factory point location table, and forming a point location vector; performing correlation calculation based on the point location vectors to obtain correlation values between any different feature point locations, and storing the correlation values to form a correlation database; according to the correlation threshold, screening out feature point locations related to the target value to be predicted from a correlation database, and forming a related point location list corresponding to the target point location; and based on the related point location list, modeling the sampling data in the preset time period by adopting a multi-dimensional time sequence algorithm, and outputting a future period prediction value corresponding to the target value to be predicted. Through the processing scheme provided by the invention, the historical sampling data is analyzed, business driving is converted into data driving, and the accuracy of modeling analysis is improved by utilizing internal association among the data.

The invention provides a semiconductor plant heat recovery system control method and device, a terminal and a storage medium, and relates to the technical field of flow and temperature control of semiconductor plant heat recovery systems. The method comprises the following steps: constructing a heat recovery automatic control model of the semiconductor plant heat recovery system, wherein the heat recovery automatic control model comprises a flow distribution mechanism model and a data model; the flow distribution mechanism model is used for conducting distribution calculation on the inlet total flow of the semiconductor plant heat recovery system, and bypass branch flow and heat exchange branch flow are obtained; the bypass branch flow and the heat exchange branch flow are input into the data model, and the bypass valve predicted opening degree and the heat exchange valve predicted opening degree are updated in real time; and based on the predicted opening degree of the bypass valve and the predicted opening degree of the heat exchange valve, the semiconductor plant heat recovery system is controlled to operate, so that the convergence temperature of the semiconductor plant heat recovery system reaches the preset convergence temperature. Accurate temperature control, valve precision driving and energy consumption optimization can be achieved.

The invention provides a tank car operation behavior analysis method based on an improved YOLO network for a semiconductor factory service system, and belongs to the field of semiconductor manufacturing, and the method specifically comprises the steps: collecting a real-time image of a tank car operation region; marking a target area corresponding to multiple view angles of each detection target in the real-time image through an improved YOLO network; generating a real-time behavior sequence according to the target area of each detection target marked by the real-time image; judging whether the real-time behavior sequence conforms to a specification or not; judging whether the operation behavior of the tank car occurs at a specified position or not by combining a mapping relation between the marked target area of each detection target and the area position of the shooting scene; and outputting a real-time analysis result corresponding to the tank car operation behavior according to the real-time behavior sequence and a judgment result of the tank car operation behavior. Through the processing scheme provided by the invention, all-around intelligent supervision of tank car operation from physical operation to flow compliance is realized, and the safety and management efficiency of a high-risk operation scene are remarkably improved.