Full-automatic microbial growth curve detection system based on deep learning

Abstract

The invention discloses a full-automatic microorganism growth curve detection system based on deep learning, which comprises an upper computer module and a microorganism growth curve detection device, and is characterized in that the microorganism growth curve detection device comprises a case, and a shaking table module, a temperature control module, a light source module and a detection module which are arranged in the case; all the modules are connected with the main controller; the shaking table module is used for bearing an incubator in the temperature control module, and four 96-hole microorganism culture plates can be placed in the incubator. The light source is located on the right side in the case, and light source wavelength can be switched and transmitted to the detection module through optical fibers. The detection module comprises a detector and an XY-axis moving device, and the detector is fixed on the XY-axis moving device. The incubator is positioned in the middle of the detector. The upper computer software reads the working information, temperature and absorbance data of each module, and displays a microbial growth curve in real time according to the absorbance data. Meanwhile, the upper computer predicts the curve through a deep learning method, and detection can be ended in advance. The detection device can realize real-time detection and prediction of the growth process of various microorganisms, realizes high-throughput growth curve measurement, simplifies the operation of experimenters, shortens the detection time, and improves the experiment efficiency and the determination accuracy.

Description

Technical Field

[0001] The present invention relates to the field of microbial detection, and in particular to a fully automatic microbial growth curve detection system based on deep learning. Background Art

[0002] With the continuous development of science and technology, people pay more and more attention to the detection and research of microorganisms. At the same time, microbial growth curve is a commonly used method for microbial detection. By measuring the growth curve, we can obtain information about the growth of microorganisms and provide quantitative growth kinetic parameters, which helps to understand the dynamic changes of cell growth and metabolism. However, the current common method of microbial growth curve detection is that the experimenter manually takes out the bacterial solution every 1 to 2 hours and uses a spectrophotometer to measure the absorbance. The entire experimental cycle is long, usually requiring the experimenter to operate continuously overnight and has high professional and technical requirements for the experimenter. The detection efficiency is low, and it is easy to cause contamination of the bacterial solution when taking out the sample bacterial solution. Summary of the invention

[0003] The purpose of this paper is to provide a fully automatic microbial growth curve detection system based on deep learning to solve the problems of small sample size, low detection efficiency and poor detection accuracy when detecting microbial samples.

[0004] In order to solve the above problems, the implementation method of this article is as follows: This article provides a fully automatic microbial growth curve detection system based on deep learning, including a host computer module and a microbial growth curve detection device.

[0005] Specifically, the fully automatic microbial growth curve detection device includes a chassis, on which is a flip-open light shield, and the chassis is provided with a light source module, a detection module, a temperature control module and a shaker module, and the light source module, the detection module, the temperature control module and the shaker module are all connected to a main controller.

[0006] Specifically, the light source module is used to provide monochromatic light required for detection under the control of the main controller, and a fiber optic transmission light source is arranged in the light source module; the temperature control module is used to stabilize the temperature in the culture area at a set temperature under the control of the main controller, and provide a temperature environment required for the growth of microorganisms; an incubator is arranged in the temperature control module, and the incubator is made of heat-conductive metal aluminum. The incubator provides four culture troughs for placing 96-well plates, and the incubator is provided with a special incubator cover, and the incubator cover corresponds one-to-one to the well positions of the 96-well plates; the detection module is used to traverse all the 96-well plates in the culture troughs to complete the detection under the control of the main controller; the shaker module is used to drive the incubator to oscillate under the control of the main controller, thereby realizing the oscillation of the sample.

[0007] Preferably, the light source module includes a shading box, an aluminum block for LED heat dissipation, a filter switching motor, and a full-spectrum LED, a filter switching wheel, an LED driving board, and an optical fiber arranged in the shading box; the filter switching wheel is provided with 5 filter slots for placing filters, and the filters in the filter slots can be disassembled and replaced; the aluminum block for LED heat dissipation and the filter switching motor are installed on both sides of the shading box; the LED driving board is used to drive the full-spectrum LED on and off; the filter switching wheel is installed on the rotating shaft of the filter switching motor, and the filter switching motor drives the filter switching wheel to realize the switching of filters, and the monochromatic light generated by the filter is transmitted by the optical fiber.

[0008] Furthermore, a fixing structure is provided in the light shielding box to fix the parts of the light source module together, wherein the full-spectrum LED and the filter slot are aligned with the center of the optical fiber.

[0009] Preferably, the detection module includes a detector and an XY axis moving device; specifically, the Y axis is mounted on the sliding bar of the X axis, and the sliding bar is fixed on the sliding bar of the Y axis; the XY axis moving device is provided with a first motor and a track for driving the Y axis to move, and a second motor and a track for driving the sliding bar to move; the detector is fixed on the sliding bar. The first motor and the second motor are connected to the main controller through a control circuit, and the main controller controls their rotation direction, wherein a first infrared sensor and a second infrared sensor are provided beside the X axis and the Y axis, wherein the first infrared sensor and the second infrared sensor are connected to the main controller, and are used to provide the initial position of the Y axis on the sliding bar of the X axis and the initial position of the sliding bar on the sliding bar of the Y axis.

[0010] Furthermore, the detector is provided with a reflector, a convex lens and a detection main board. The other end of the optical fiber is fixed to the reflector and the convex lens through an L-shaped frame upper cover and an L-shaped frame base; the optical fiber and the convex lens are respectively aligned with two surfaces of the reflector; the monochromatic light transmitted by the optical fiber is converged into the sample via the reflector and the convex lens.

[0011] Furthermore, a detection circuit board with a photoelectric sensor detection circuit is provided directly below the convex lens, and the photoelectric detection circuit is connected to the main controller. The main controller can read the detection data in real time, and the main controller will send the data to the host computer through the serial port; the detection circuit board is installed under the detection cover, and the detection cover is fixed to the detection bottom plate; the L-shaped frame base is fixed on the detection cover.

[0012] Preferably, the temperature control module is composed of an incubator, a heating film, a temperature sensor, a first refrigerator and a second refrigerator. The heating film and the temperature sensor are arranged in the incubator; the incubator is provided with an incubator cover, a locking block and an incubator bottom plate; the heating film and the incubator cover are highly coupled in structure; the incubator bottom plate provides 4 96-well plate slots, and the incubator cover and the heating film correspond to each culture well of the 96-well plate. The structure of the incubator cover and the heating film ensures that the temperature of each detection well of the 96-well plate is highly uniform, and avoids interference between detections of different wells.

[0013] Furthermore, the heating film, the temperature sensor, the first refrigerator and the second refrigerator are connected to the main controller through a temperature control circuit. The host computer sends the temperature value set by the main controller through the serial port, and the main controller reads the temperature value of the temperature sensor through the temperature control circuit, and then the main controller controls the heating and the operation of the first refrigerator and the second refrigerator through the temperature control circuit. When the temperature read by the main controller is less than the received set value, the heating film is operated through the temperature control circuit, and the first refrigerator and the second refrigerator are turned off; when the temperature read by the main controller is greater than the received set value, the heating film is turned off through the temperature control circuit, and the first refrigerator and the second refrigerator are operated.

[0014] Preferably, the rocking bed module includes a rocking bed motor, a rocking bed partition, an eccentric shaft, four synchronous eccentric shafts, an infrared sensor, a rocking bed bottom plate and six support frames. The rocking bed motor is installed under the rocking bed partition; the eccentric shaft is installed on the rotating shaft of the rocking bed motor and is separated by the rocking bed partition; the synchronous eccentric shaft has the same wheelbase and length as the eccentric shaft and is fixed on the rocking bed partition; the rocking bed bottom plate is provided with holes corresponding to the eccentric shaft and the four synchronous eccentric shafts; the six support frames are fixed to the rocking bed bottom plate, specifically four of which are located on the right side of the rocking bed bottom plate and two are located on the left side of the rocking bed bottom plate.

[0015] Furthermore, the rocking bed motor and infrared sensor are connected to the main controller through a control circuit. The main controller can control the rotation speed of the rocking bed motor through the control circuit, and then the eccentric shaft is driven by the rotating shaft of the rocking bed motor. Because of the eccentric shaft, the synchronous eccentric shaft is fixed together with the rocking bed base plate, and the synchronous eccentric shaft can rotate together. Here, the synchronous eccentric shaft only plays a supporting role, and the rocking bed base plate will follow its rotation and oscillate in a small range.

[0016] Furthermore, the incubator is fixed on the support frame and moves along with the movement of the shaker bottom plate, thereby realizing a small-scale oscillation of the sample in the incubator.

[0017] Preferably, the interface of the host computer mainly includes parameter setting items of "detection wavelength", "number of orifice plates", "temperature", "detection frequency", "rotation speed", "detection duration", and "prediction model", "set parameters", "start detection", "stop detection" buttons and "curve display" and "device status information display" areas.

[0018] Furthermore, the detection data is continuously generated and sent to the host computer by the main controller. On the one hand, it is displayed in real time in the "curve display" area of ​​the host computer interface, and on the other hand, the host computer pre-processes the detection data and sends it to the database.

[0019] Furthermore, after the data arrives at the database, the time window is set to T, and the detection data at the current moment and the detection data at the previous T-1 moments are used as the input of the model selected for model selection for prediction, and the highly accurate prediction results of the model are displayed on the growth curve of the corresponding sample.

[0020] Compared with the prior art, the present invention has the following advantages: A fully automatic microbial growth curve detection system based on deep learning integrates microbial detection and microbial cultivation, and can realize real-time detection and prediction of the microbial growth process. The detection device realizes the oscillation cultivation of microorganisms through the temperature control module and the shaking table module inside the chassis, ensuring that the microorganisms are evenly distributed in the sample during the cultivation process. The light source module and the detection module realize the measurement of the absorbance of the sample during the cultivation process. The main controller reads the change of light intensity by reading the detection mainboard and sends it to the host computer. The host computer 80 can set the initialization parameters of the microbial growth curve detection device, display the detection data in the form of a curve, and use the detection data and the deep learning method to predict the curve. The fully automatic microbial growth curve detection system based on deep learning can realize the detection of 384 samples in 4 96-well plates at most, realize the detection of high-throughput microbial growth curves, simplify the operation of the experiment, reduce the labor intensity of the experimenter, and improve the experimental efficiency and the accuracy of microbial detection.

[0021] In order to make the above-mentioned purposes, features and advantages of this article more obvious and easy to understand, specific implementation cases are given below for detailed description, and illustrations are attached. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the overall fully automatic microbial growth curve detection system based on deep learning provided in the embodiments of this article.

[0023] Figure 2 Schematic diagram of a microbial growth curve detection device of a fully automatic microbial growth curve detection system based on deep learning provided in an embodiment of this article.

[0024] Figure 3 It is a schematic diagram of the internal structure of the microbial growth curve detection device.

[0025] Figure 4 This is a control logic diagram of a fully automatic microbial growth curve detection system based on deep learning provided in an embodiment of this article.

[0026] Figure 5 It is a schematic diagram of the structural decomposition of the light source module of the microbial growth curve detection device.

[0027] Figure 6 It is a schematic diagram of the structural decomposition of the detection module of the microbial growth curve detection device.

[0028] Figure 7 It is a schematic diagram of the structural decomposition of the temperature control module of the microbial growth curve detection device.

[0029] Figure 8 It is a schematic diagram of the structure of a shaker module of a microbial growth curve detection device.

[0030] Fig. 9 It is a top view of the shaker module of the microbial growth curve detection device.

[0031] Fig.10 It is the operation interface of the host computer.

[0032] Fig.11 This is a flowchart of using deep learning methods to make curve predictions Implementation

[0033] The technical solution in this embodiment will be described in detail below in conjunction with the drawings in the embodiments of this article.

[0034] It should be noted that the terms "first", "second", etc. in the specification and claims of this document and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0035] The embodiment of this article provides a fully automatic microbial growth curve detection system based on deep learning, such as Figure 1 As shown, it includes a microbial growth curve detection device 10 and a host computer 80.

[0036] Specifically, the fully automatic microbial growth curve detection device includes a chassis 101, on which a flip-openable light shield 102 is provided, and inside the chassis are provided a light source module 20, a detection module 30, a temperature control module 40 and a shaking table module 50. The light source module 20, the detection module 30, the temperature control module 40 and the shaking table module 50 are all connected to a main controller 70.

[0037] The light source module 20 is used to provide monochromatic light required for detection under the control of the main controller 70, and an optical fiber 27 is arranged in the light source module 20 to transmit the light source; the temperature control module 40 is used to stabilize the temperature in the culture area at a set temperature under the control of the main controller 70, and provide a temperature environment required for the growth of microorganisms; the temperature control module 30 is provided with an incubator 43, and the incubator 43 is made of heat-conductive metal aluminum. The incubator 43 provides a culture tank 44 for placing four 96-well plates, and the incubator 43 is provided with a special incubator cover 431, and the incubator cover 431 corresponds to the well positions of the 96-well plates one by one; the detection module 30 is used to traverse all the 96-well plates in the culture tank 44 to complete the detection under the control of the main controller 70; the shaker module 50 is used to drive the incubator 43 to oscillate under the control of the main controller 70, thereby realizing the oscillation of the sample.

[0038] In the embodiment of this specification, the main controller 70 may be a microprocessor (MCU), a single-chip microcomputer, etc., or a control board integrated with the main controller, etc. Figure 3 As shown, the light source module 20, the detection module 30, the temperature control module 40 and the rocking bed module 50 are all connected to the main controller 70, which means that the light source module 20, the detection module 30, the temperature control module 40 and the rocking bed module 50 are connected to the main controller 70 through wired or wireless communication technology.

[0039] like Figure 4 As shown, in the embodiment of the present specification, the light source module 20 includes a light shielding box 29, an aluminum block 22 for LED heat dissipation, a filter switching motor 26, and a full-spectrum LED 23, a filter switching wheel 24, an LED driving board 25, and an optical fiber 27 arranged in the light shielding box 29; the filter switching wheel 24 is provided with 5 filter slots 28 for placing filters, and the filters in the filter slots 28 can be disassembled and replaced; the aluminum block 22 for LED heat dissipation and the filter switching motor 26 are installed on both sides of the light shielding box 29; the LED driving board is used to drive the full-spectrum LED 23 to turn on and off; the filter switching wheel 24 is installed on the rotating shaft of the filter switching motor 26, and the filter switching motor 26 drives the filter switching wheel 24 to realize the switching of the filter, and the monochromatic light generated by the filter is transmitted by the optical fiber 27.

[0040] Specifically, in this embodiment, a fixing structure is provided in the light shielding box 29 to fix the parts of the light source module 20 together, wherein the full-spectrum LED 23 and the filter slot 28 are aligned with the center of the optical fiber 27 .

[0041] like Figure 5As shown, in the embodiment of this specification, the detection module 30 includes a detector 32 and an XY axis moving device 31; specifically, the Y axis is mounted on the slide bar 313 of the X axis, and the slider 316 is fixed on the slide bar 315 of the Y axis; the XY axis moving device 31 is provided with a first motor 311 and a crawler 312 to drive the Y axis to move, and a second motor (not shown in the figure) and a crawler 317 to drive the slider 316 to move; the detector 32 is fixed on the slider 316. The first motor 311 and the second motor (not shown in the figure) are connected to the main controller 70 through a control circuit, and the rotation direction thereof is controlled by the main controller 70, wherein a first infrared sensor 314 and a second infrared sensor (not shown in the figure) are provided next to the X axis and the Y axis, wherein the first infrared sensor 314 and the second infrared sensor (not shown in the figure) are connected to the main controller 70, and are used to provide the initial position of the Y axis on the slide bar 313 of the X axis and the slider on the slide bar 315 of the Y axis.

[0042] Specifically, the detector 32 is provided with a reflector 322, a convex lens 323 and a detection main board 326. The other end of the optical fiber 27 is fixed to the reflector 322 and the convex lens 323 through an L-shaped frame upper cover 321 and an L-shaped frame base 324; the optical fiber 27 and the convex lens 323 are respectively aligned with two surfaces of the reflector 322; the monochromatic light transmitted by the optical fiber 27 is converged into the sample via the reflector 323 and the convex lens 323.

[0043] Specifically, a detection circuit board 326 with a photoelectric sensor detection circuit is provided directly below the convex lens 323. The photoelectric detection circuit is connected to the main controller 70. The main controller 70 can read the detection data in real time, and the main controller 70 will send the data to the host computer 80 through the serial port; the detection circuit board 326 is installed under the detection cover 325, and the detection cover 325 is fixed to the detection bottom plate 327; the L-shaped frame base 324 is fixed on the detection cover 325.

[0044] like Figure 5 As shown, in this embodiment, the temperature control module 40 is composed of an incubator 43 , a heating film 46 , a temperature sensor 45 , a first refrigerator 41 and a second refrigerator 42 .

[0045] Specifically, the heating film 46 and the temperature sensor 45 are arranged in the incubator 43; the incubator 43 is provided with an incubator cover 431, a locking block 432, and an incubator bottom plate 433; the heating film 46 and the incubator cover 431 are highly coupled in structure; the incubator bottom plate 433 provides four 96-well plate slots, and the incubator cover 431 and the heating film 46 correspond to each culture well of the 96-well plate. The structure of the incubator cover 431 and the heating film 46 ensures that the temperature of each detection well of the 96-well plate is highly uniform, and avoids interference between detections of different wells.

[0046] Specifically, the heating film 46, the temperature sensor 45, the first refrigerator 41 and the second refrigerator 42 are connected to the main controller 70 through the temperature control circuit. The upper computer 80 sends the temperature value set by the main controller 70 through the serial port, and the main controller 70 reads the temperature value of the temperature sensor 45 through the temperature control circuit, and then the main controller 70 controls the operation of the heating film 46 and the first refrigerator 41 and the second refrigerator 42 through the temperature control circuit. When the temperature read by the main controller 70 is less than the received set value, the heating film 46 is operated through the temperature control circuit, and the first refrigerator 41 and the second refrigerator 42 are turned off; when the temperature read by the main controller 70 is greater than the received set value, the heating film 46 is turned off through the temperature control circuit, and the first refrigerator 41 and the second refrigerator 42 are operated.

[0047] like Figure 7 , Figure 8 As shown, in this embodiment, the rocking bed module 50 includes a rocking bed motor 57, a rocking bed partition 53, an eccentric shaft 51, four synchronous eccentric shafts 54, an infrared sensor 52, a rocking bed bottom plate 56 and six support frames 55. The rocking bed motor 57 is installed under the rocking bed partition 53; the eccentric shaft is installed on the rotating shaft of the rocking bed motor 57 and is separated by the rocking bed partition 53; the synchronous eccentric shaft 54 ​​has the same wheelbase and length as the eccentric shaft 51, and is fixed on the rocking bed partition 56; the rocking bed bottom plate 56 is provided with holes corresponding to the eccentric shaft 51 and the four synchronous eccentric shafts 54 for installation; the six support frames 55 are fixed to the rocking bed bottom plate 56, specifically, four are located on the right side of the rocking bed bottom plate 56, and two are located on the left side of the rocking bed bottom plate 56.

[0048] Specifically, the rocking bed motor 57 and the infrared sensor 52 are connected to the main controller 70 through a control circuit. The main controller 70 can control the rotation speed of the rocking bed motor 57 through the control circuit, and then the eccentric shaft 54 ​​is driven by the rotating shaft of the rocking bed motor 57. Since the eccentric shaft 54, the synchronous eccentric shaft 54 ​​is fixed together with the rocking bed base plate 56, the synchronous eccentric shaft 54 ​​can rotate together. Here, the synchronous eccentric shaft 54 ​​only plays a supporting role, and the rocking bed base plate 56 will follow its rotation to oscillate in a small range.

[0049] Specifically, the incubator 43 is fixed on the support frame 55 and moves along with the movement of the shaker bottom plate 56 , thereby realizing a small-scale oscillation of the sample in the incubator 43 .

[0050] like Fig. 9 , Fig.10 As shown, in this embodiment, the upper computer interface 840 mainly includes parameter setting items of "detection wavelength", "number of orifice plates", "temperature", "detection frequency", "rotation speed", "detection time", and "prediction model", "set parameters", "start detection", "stop detection" buttons and "curve display", "device status information display" areas.

[0051] Specifically, the detection data 12 is continuously generated and sent to the host computer 80 by the main controller 70. On the one hand, it is displayed in real time in the "curve display" area of ​​the host computer interface 840, and on the other hand, the host computer 80 pre-processes 810 the detection data and sends it to the database 811.

[0052] Specifically, after the data arrives at the database 811, the time window is set to T, and the detection data at the current moment and the detection data at the previous T-1 moments are used as inputs of the model 821 selected by the model selection 820 for prediction, and the high-accuracy prediction results of the model are displayed on the growth curve of the corresponding sample.

[0053] In order to more clearly illustrate the fully automatic microbial growth curve detection device in this embodiment, its working process is as follows: (1) After the device is powered on, each module inside the chassis will perform a self-check, the position of the XY axis moving device 31 will be reset, the eccentric shaft 51 will be reset, the filter switching wheel 24 will be reset, and the host computer will display the device status information.

[0054] (2) The user first sets the "detection wavelength" used by the required light source module 20, the "number of 96 well plates" to be detected, the "temperature", "detection duration", and "detection frequency" to be maintained by the temperature control module 40, the "speed" of the shaker motor 57 of the shaker module 50, and the required prediction model. Then click the "Set Parameters" button to save the set parameters.

[0055] (3) If there is no fault, wait until the temperature in the cabinet is maintained at the set temperature, open the light shielding cover 11, and place the prepared 96-well plate in the incubator 43.

[0056] (4) The user clicks the "Start Detection" button on the host computer 80 interface and the detection begins.

[0057] (5) After the detection starts, the shaking table module 50 drives the incubator 43 to oscillate in a small range, and then the XY axis moving device 31 drives the detector 32 to cover the detection area of ​​all 96-well plates for detection. After the detection of all plate positions is completed each time, the shaking table module 50 is controlled to drive the incubator 43 to oscillate in a small range, and then repeat the detection until the detection is completed. During this period, the host computer uses the pre-processed detection data at the current time T and the previous T-1 time as the input of the prediction model to predict the subsequent data. The host computer displays the detection data and the predicted data in real time in the "curve display" area.

[0058] (6) The user can click the "Stop Detection" button in advance or wait for the detection time to complete according to the predicted curve.

[0059] (7) After stopping the test, the user takes out the 96-well plate placed in the incubator 43.

[0060] Furthermore, the prediction model is a network model of the type such as a trained recurrent neural network (RNN) or a long short-term memory network (LSTM).

[0061] Furthermore, the data preprocessing mainly includes: data normalization or standardization, data format conversion, and data size matching. The detection data is normalized or standardized to make it range between 0 and 1 or have a standard normal distribution to avoid the impact of scale differences on the model, and the detection data is converted into a Tensor format suitable for the model, and the data is ensured to match the input size of the model.

[0062] The above-mentioned fully automatic microbial growth curve detection system based on deep learning integrates microbial curve detection, microbial curve prediction, and microbial cultivation, and can realize real-time detection and prediction of the microbial growth process. The detection system is divided into a microbial growth curve detection device 10 and a host computer 80. The microbial growth curve detection device realizes the oscillation cultivation of microorganisms through the temperature control module 40 and the shaking table module 50 inside the chassis 10, ensuring that the microorganisms are evenly distributed in the sample during the cultivation process. The light source module 20 and the detection module 30 realize the measurement of the absorbance of the sample during the cultivation process. The main controller 70 reads the change in light intensity by reading the detection mainboard 326 and sends it to the host computer 80. The host computer 80 can set the initialization parameters of the microbial growth curve detection device, display the detection data in the form of a curve, and use the detection data and the deep learning method to predict the curve. This deep learning-based fully automatic microbial growth curve detection system can detect up to 384 samples in 4 96-well plates, achieving high-throughput microbial growth curve detection, simplifying experimental operations, reducing the labor intensity of experimenters, and improving experimental efficiency and the accuracy of microbial detection.

[0063] This article uses specific embodiments to illustrate the principles and implementation methods of this article. This narrative style of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation methods that can be understood by those skilled in the art.

Claims

1. A fully automatic microbial growth curve detection system based on deep learning, including a host computer module and a microbial growth curve detection device.

2. According to claim 1, a fully automatic microbial growth curve detection system based on deep learning is characterized in that The microbial growth curve detection device includes a chassis and a shaking table module, a temperature control module, a light source module, and a detection module arranged in the chassis, and is characterized in that: the light source module is used to provide monochromatic light required for detection under the control of the main controller, and a fiber optic transmission light source is arranged in the light source module; the temperature control module is used to stabilize the temperature in the culture area at a set temperature under the control of the main controller, and provide a temperature environment required for microbial growth; the temperature control module is provided with an incubator, the incubator is made of heat-conductive metal aluminum, and the incubator provides 4 culture tanks for placing 96-well plates, the incubator is provided with a special incubator cover, and the incubator cover corresponds to the well positions of the 96-well plates one by one; the detection module is used to traverse all the 96-well plates in the culture tanks to complete the detection under the control of the main controller; the shaking table module is used to drive the incubator to oscillate under the control of the main controller, thereby realizing the oscillation of the sample.

3. According to claim 2, a fully automatic microbial growth curve detection system based on deep learning is characterized in that: The light source module includes a shading box, an aluminum block for LED heat dissipation, a filter switching motor, and a full-spectrum LED, a filter switching wheel, an LED driving board, and an optical fiber arranged in the shading box; the filter switching wheel is provided with 5 filter slots for placing filters, and the filters in the filter slots can be disassembled and replaced; the aluminum block for LED heat dissipation and the filter switching motor are installed on both sides of the shading box; the LED driving board is used to drive the full-spectrum LED on and off; the filter switching wheel is installed on the rotating shaft of the filter switching motor, and the filter switching motor drives the filter switching wheel to realize the switching of filters, and the monochromatic light generated by the filter is transmitted by the optical fiber.

4. According to claim 2, a fully automatic microbial growth curve detection system based on deep learning is characterized in that: A fixing structure is arranged in the light shielding box to fix the parts of the light source module together, wherein the full-spectrum LED and the filter slot are aligned with the center of the optical fiber.

5. According to claim 2, a fully automatic microbial growth curve detection system based on deep learning is characterized in that: The detection module includes a detector and an XY axis moving device; specifically, the Y axis is installed on the sliding bar of the X axis, and the sliding bar is fixed on the sliding bar of the Y axis; the XY axis moving device is provided with a first motor and a track to drive the Y axis to move, and a second motor and a track to drive the sliding bar to move; the detector is fixed on the sliding bar. The first motor and the second motor are connected to the main controller through a control circuit, and the main controller controls their rotation direction, wherein a first infrared sensor and a second infrared sensor are provided beside the X axis and the Y axis, wherein the first infrared sensor and the second infrared sensor are connected to the main controller, and are used to provide the initial position of the Y axis on the sliding bar of the X axis and the sliding bar on the sliding bar of the Y axis.

6. According to claim 5, a fully automatic microbial growth curve detection system based on deep learning is characterized in that: The detector is provided with a reflector, a convex lens and a detection main board. The other end of the optical fiber is fixed with the reflector and the convex lens through an L-shaped frame upper cover and an L-shaped frame base; the optical fiber and the convex lens are respectively aligned with two surfaces of the reflector.

7. According to claim 5, a fully automatic microbial growth curve detection system based on deep learning is characterized in that: A detection circuit board with a photoelectric sensor detection circuit is provided directly below the convex lens. The photoelectric detection circuit is connected to the main controller. The main controller can read the detection data in real time, and the main controller will send the data to the host computer through the serial port; the detection circuit board is installed under the detection cover, and the detection cover is fixed to the detection bottom plate; the L-shaped frame base is fixed on the detection cover.

8. According to claim 2, a fully automatic microbial growth curve detection system based on deep learning is characterized in that: The temperature control module is composed of an incubator, a heating film, a temperature sensor, a first refrigerator and a second refrigerator. The heating film and the temperature sensor are arranged in the incubator; the incubator is provided with an incubator cover, a locking block and an incubator bottom plate; the heating film and the incubator cover are highly coupled in structure; the incubator bottom plate provides 4 96-well plate slots, and the incubator cover and the heating film correspond to each culture well of the 96-well plate.

9. A fully automatic microbial growth curve detection system based on deep learning according to claim 8, characterized in that: The heating film, the temperature sensor, the first refrigerator and the second refrigerator are connected to the main controller through the temperature control circuit. The upper computer sends the temperature value set by the main controller through the serial port, and the main controller reads the temperature value of the temperature sensor through the temperature control circuit, and then the main controller controls the heating and the operation of the first refrigerator and the second refrigerator through the temperature control circuit. When the temperature read by the main controller is less than the received set value, the heating film is operated through the temperature control circuit, and the first refrigerator and the second refrigerator are turned off; when the temperature read by the main controller is greater than the received set value, the heating film is turned off through the temperature control circuit, and the first refrigerator and the second refrigerator are operated.

10. The fully automatic microbial growth curve detection system based on deep learning according to claim 2, characterized in that: The rocking bed module includes a rocking bed motor, a rocking bed partition, an eccentric shaft, four synchronous eccentric shafts, an infrared sensor, a rocking bed bottom plate and six support frames. The rocking bed motor is installed under the rocking bed partition; the eccentric shaft is installed on the rotating shaft of the rocking bed motor and is separated by the rocking bed partition; the synchronous eccentric shaft has the same wheelbase and length as the eccentric shaft and is fixed on the rocking bed partition; the rocking bed bottom plate is provided with holes corresponding to the eccentric shaft and the four synchronous eccentric shafts; the six support frames are fixed to the rocking bed bottom plate, specifically four of which are located on the right side of the rocking bed bottom plate and two are located on the left side of the rocking bed bottom plate.

11. A fully automatic microbial growth curve detection system based on deep learning according to claim 10, characterized in that: The rocking bed motor and infrared sensor are connected to the main controller through a control circuit. The main controller can control the rotation speed of the rocking bed motor through the control circuit, and then the eccentric shaft is driven by the rotating shaft of the rocking bed motor. Because of the eccentric shaft, the synchronous eccentric shaft is fixed together with the rocking bed base plate, and the synchronous eccentric shaft can rotate together. Here, the synchronous eccentric shaft only plays a supporting role, and the rocking bed base plate will follow its rotation and oscillate in a small range.

12. The fully automatic microbial growth curve detection system based on deep learning according to claim 10, characterized in that: The incubator is fixed on the support frame and moves along with the movement of the shaker bottom plate.

13. The fully automatic microbial growth curve detection system based on deep learning according to claim 1, characterized in that: The host computer module is used for parameter setting and real-time prediction curve display of the microbial growth curve detection device. The real-time prediction curve is composed of real-time detection values ​​and prediction values.

Images