Ai Mineral Fluospar CCD Color Sorter Barite Sorting Machine

• Type: Optical Ore Sorter
• Voltage: 220V
• Weight: 1.5ton
• Model NO.: MAI-D-4
• Material: Stainless Steel
• Material Feature: Heat Resistant
• Certification: CE, ISO9001:2015
• Energy Saving: Energy Saving
• Warranty: 1 Year
• Color: Grey
• Customized: Customized
• Condition: New
• After-sales Service: Provided
• Transport Package: Wooden Box
• Specification: 4676*2012*1690
• Trademark: MINGDE
• Origin: China
• Production Capacity: 2000 Sets/Year

Product Description

Product Description

Mingde Optoelectronics' AI sorting machine innovatively integrates deep convolutional neural networks (CNN) with visible light sorting technology. Through its independently developed multi-dimensional feature extraction algorithm, it achieves precise identification and efficient sorting of complex ores. This device adopts the unique local connection, weight sharing and multi-convolutional kernel architecture of CNN. During the training process, it independently builds a multi-dimensional feature database of materials, and the sorting accuracy significantly exceeds that of traditional photoelectric sorting machine.

Working principle

1. High-precision multispectral/hyperspectral sensing: The equipment is equipped with advanced optical sensors that can accurately capture subtle spectral feature differences on the surface and inside of the ore (such as color, texture, mineral composition, impurity distribution, etc.), with a resolution far exceeding traditional manual or mechanical sorting methods.
2. AI-driven intelligent recognition algorithm: The core algorithm based on deep learning can process massive spectral data in real time and accurately establish a mapping model between ore characteristics and target minerals/impurities. The system has strong self-learning and adaptive capabilities and can cope with natural fluctuations in ore properties.
3. High-speed real-time processing and precise blowing: The powerful data processing platform ensures that when the ore passes at high speed (the processing capacity can reach tens to hundreds of tons/hour), the identification and decision-making are completed instantly, and the high-pressure gas valve array is driven to spray and separate the target particles with millisecond-level precision.
4. Modular and customized design: The equipment adopts a modular design, and the core optical module, sorting execution module, control module, etc. can be flexibly configured according to different ore characteristics (particle size, density, sorting requirements) to meet the diverse needs from coarse-grained pre-throwing to fine-grained selection.

Technical Parameters

Product	Mineral particle size(mm)	Capacity(t/h)	Air pressure(mpa)	Sorting accuracy(%)	Optimal take-out ratio	Power(kw)	Dimension(mm)	Weight(kg)
MAI-D-2	5-10	3-4	0.55	96	10:1	3	2945*1390*1640	1000
	10-30	5-6	0.6
	30-60	15-23	0.65
MAI-D-4	5-10	6-8	0.55	96	10:1	5	4680*2015*1700	1500
	10-30	10-12	0.6
	30-60	30-45	0.65
MAI-D-6	5-10	9-12	0.55	96	10:1	7	5200*2502*1925	3500
	10-30	15-18	0.6
	30-60	45-68	0.65
MAI-TD-6	30-80	45-80	0.6	96	10:1	10	6790*2502*2190	6000
	80-150	150-200	0.65
MAI-S-4	5-10	7-9	0.55	99.8	20:1	9	5155*3430*2680	3500
	10-30	12-14	0.6
	30-60	35-50	0.65

Technical Advantages:

1,Empowered by AI Deep Convolutional Neural Networks

The CNN artificial intelligence algorithm is applied to the field of optoelectronic sorting, and the intelligent analysis of material image features is achieved through a multi-layer neural network architecture, breaking through the bottleneck of traditional color sorting technology.

2,Multi-dimensional Stereoscopic Feature Recognition

The innovatively developed AI photoelectric sorting technology can automatically extract multi-dimensional features such as texture, shape, color, quality and luster of materials, greatly expanding its application scenarios and meeting the personalized sorting needs of different sorting scenarios and material types.

3,Real-time Dynamic Acceleration Technology

By adopting a collaborative optimization scheme of model compression and hardware, the inference speed of CNN is increased to the millisecond level, effectively guaranteeing the requirements of industrial-grade real-time sorting.

4,Small sample learning system

Integrating the transfer learning framework with industrial-grade image enhancement technology, it can still reach a high recognition accuracy rate in scenarios with a sample less than a thousand cases.

5,High-Performance Computing Platform

It is equipped with a gigabit Ethernet industrial camera and a multi-GPU parallel computing platform. The computing platform adopts CNN for material type discrimination technology, which can accurately analyze the small subtle surface features of materials.

Ores can be sorted

Certifications

FAQ

1.Can we visit your factory?

2.How long about the machine guarantee period?

Yes,our factory is located in Hefei city,Anhui Province, 2.5 hour bullet train distance from Shanghai,warmly welcome your visiting.

One year. And we supply lifelong software upgrade services for our customers.

3.Can you test samples to check sorting effect?

Yes,sample sending for testing is welcomed.