Manual trimming auxiliary program of 5g filter based on AI
Nanjing XingHuo Technology Co., Ltd. (hereinafter referred to as “XingHuo technology”) recently released a manual trimming auxiliary program for 5g filter based on AI, which can accurately judge the trimming site, give the shortest trimming path by AI, plan and guide manual trimming, and greatly improve the efficiency.
The trimming strategy of microwave devices in the cloud combines artificial intelligence algorithm with traditional filter trimming experience; it has the ability of self-learning and self evolution, which makes the trimming speed faster and faster.
Users create trimming tasks based on local clients:
➤ it can be created by date or by product type (batch);
➤ select the products to be trimmed, such as dielectric filter, cavity filter, duplexer, multiplexer, etc;
➤ input the index of the product to be trimmed: center frequency, passband, return loss, stopband attenuation, etc.
Show video
Based on the topology of duplexer and multiplexer, its resonance and coupling are extracted, and its virtual trimming environment is established. The duplexer simulates 37 trimming actions, and the multiplexer simulates 51 trimming actions. The AI trimming assistant can efficiently trim the duplexer to meet the index requirements, and the passing rate is 100%.
The manual trimming auxiliary program of 5g filter based on artificial intelligence developed by XingHuo technology perfectly combines the multi-disciplinary theories of electromagnetism, mathematics, ceramic technology and long-term manual trimming experience, which can bring more accurate and effective guidance and effectively guarantee the large-scale production of 5g filter.
Technology leads change XingHuo launched manual trimming auxiliary program of 5g filter based on Artificial Intelligence
Following the world’s first 5g filter intelligent trimming robot, Nanjing XingHuo Technology Co., Ltd. (hereinafter referred to as XingHuo technology) released the manual trimming auxiliary program of 5g filter based on AI cloud again, promoting the large-scale manual trimming technology to a new level.
Unique advantages
➤ given any state, accurately judge the trimming site
➤ AI gives the shortest trimming path to guide manual trimming and greatly improve efficiency
Technical challenges
Ceramic dielectric filter has great advantages in 5g field (small size, light weight, high attenuation, low insertion loss, good temperature stability), and has gradually replaced metal cavity filter as the mainstream scheme.
Compared with the traditional filter, ceramic filter has higher requirements for trimming. For example, the current mainstream 10 order ceramic filter has complex structure and many trimming actions. And there is irreversible operation in the whole trimming process. If manual trimming is adopted, the production efficiency and product qualification rate will be greatly reduced.
Therefore, trimming is very important in the production process of ceramic dielectric filter. In order to eliminate the influence of human factors on trimming and improve production efficiency and automation level, an auxiliary program is needed to plan and guide manual trimming.
Key algorithms
XingHuo technology uses AI cloud to build a virtual simulation environment. The most advanced couple matrix technology is used to extract the parameters of the filter, and the deep learning technology is used to transform the frequency response of the filter into a hidden state in an abstract state space through a representation function, Iterate through the previous hidden state and the assumed next action.
The first layer of the software is graphical interface, which is used for visual operation. Users can enter the second layer of functional modules through the graphical interface, such as setting the order number of filters, topology, data acquisition, autonomous training, etc. the execution status information of each module will also be reflected in the first layer of graphical interface. The third layer is data management, which is used to store the structure of neural network, the corresponding trimming strategy and other information.
The effect of trimming
concluding remarks
In order to ensure the large-scale production of ceramic dielectric filter, the improvement of trimming technology is urgent. At present, only a few companies have automatic trimming equipment, and it takes a lot of time and cost to train experienced trimming personnel, which greatly restricts the large-scale production.
XingHuo technology takes advantage of the situation. The manual trimming auxiliary program based on artificial intelligence and the AI cloud independently developed can effectively solve customers’ worries.
