A control method and system for stabilizing Taylor cone angle

Abstract

The present invention provides a control method and system for stabilizing the Taylor cone angle, the method comprising: controlling the particle-containing liquid to be sprayed from a metal nozzle according to a preset flow rate and a preset voltage; obtaining a first prediction factor according to the transmitted light intensity of the particle-containing liquid flowing through a conduit and a trained first neural network; obtaining an image of the Taylor cone angle below the metal nozzle to obtain the current Taylor cone angle; obtaining a second prediction factor according to the current Taylor cone angle and a target Taylor cone angle; and controlling the voltage of a high-voltage generator using a PID controller based on a BP neural network to stabilize the Taylor cone angle. The method uses the first prediction factor as the input of the BP neural network and combines the PID controller based on the BP neural network to control the output voltage of the high-voltage generator, so that the Taylor cone angle can be stabilized, the stability of the spray solution at the metal nozzle can be ensured, and the problem of uneven microsphere particle size or uneven microcapsule shell thickness can be avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of control and regulation systems, and in particular to a control method and system for stabilizing a Taylor cone angle. Background Art

[0002] The high-voltage electrostatic spray microsphere / microcapsule preparation method uses a high-voltage electrostatic generator to charge the sprayed droplets for the preparation of microspheres / microcapsules. The high-voltage electrostatic spray microsphere / microcapsule preparation method has been widely used in drug delivery, food industry, material science and other fields due to its unique encapsulation performance and controllable microsphere / microcapsule size.

[0003] The general process of the high-voltage electrostatic spray microsphere / microcapsule preparation method is as follows: the positive electrode of the high-voltage generator is connected to the metal nozzle, the negative electrode is connected to the equipotential ring, and the metal nozzle is suspended at a certain distance above the equipotential ring. When the power is turned on during preparation, a high-voltage electric field is generated between the two electrodes. When the syringe pump pushes the solution out at a constant speed, the electric field force overcomes the inherent viscosity and surface tension of the solution, so that microspheres of a certain particle size are sprayed into the curing liquid and quickly solidified. When a coaxial double nozzle is used, the metal nozzle connects two different solutions, and a microsphere capsule with a double-layer structure will be formed under the action of the electric field force.

[0004] Microspheres / microcapsules prepared by high-voltage electrostatic spraying have shown great application potential in many fields due to their high efficiency and controllable characteristics. However, in the actual preparation process, uneven particle size of microspheres or uneven shell thickness of microcapsules often occur, which limits their application. Summary of the invention

[0005] In view of the above defects of the prior art, the present invention provides a control method and system for stabilizing the Taylor cone angle to solve the technical problem of uneven particle size of microspheres or uneven thickness of microcapsule shells.

[0006] To achieve the above-mentioned purpose and other related purposes, the present invention provides a control method for stabilizing the Taylor cone angle, comprising: controlling the particle-containing liquid to be sprayed from a metal nozzle according to a preset flow rate and a preset voltage; obtaining the transmitted light intensity of the particle-containing liquid flowing through a conduit, and inputting the transmitted light intensity into a trained first neural network to obtain a first prediction factor; obtaining an image of the Taylor cone angle below the metal nozzle, and processing the image to obtain a current Taylor cone angle; obtaining a second prediction factor based on the current Taylor cone angle and a target Taylor cone angle; and controlling the output voltage of a high-voltage generator using a PID controller based on a BP neural network to stabilize the Taylor cone angle; wherein the input quantity of the BP neural network includes the target Taylor cone angle, the first prediction factor, the second prediction factor, and the control signal output by the PID controller.

[0007] In one embodiment of the present invention, the preset flow rate and the preset voltage are obtained according to the following steps: a plurality of scanning points are obtained according to the flow rate interval and the voltage interval; a stable jet injection interval of the particle-free liquid is obtained according to the plurality of scanning points; and the preset flow rate and the preset voltage are obtained according to the flow rate and voltage corresponding to the center point of the stable jet injection interval.

[0008] In one embodiment of the present invention, a stable jet injection interval of a particle-free liquid is obtained based on a plurality of the scanning points, including: controlling the particle-free liquid to be ejected from the metal nozzle according to the flow rate and voltage corresponding to each of the scanning points; for any scanning point: acquiring an image of the Taylor cone angle below the metal nozzle, and processing the image to obtain the Taylor cone angle; when the change amplitude or change rate of the Taylor cone angle within a preset time is less than or equal to a preset value, the scanning point is a Class I scanning point; when the change amplitude or change rate of the Taylor cone angle within a preset time is greater than a preset value, the scanning point is a Class II scanning point; based on the Class I scanning point and the Class II scanning point, the stable jet injection interval is obtained.

[0009] In one embodiment of the present invention, the sampling frequency of the transmitted light intensity is N times the sampling frequency of the image, where N is a positive integer; the number of input layer nodes of the first neural network is N, the number of intermediate layer nodes is 60, and the number of output layer nodes is 1, and N consecutive transmitted light intensities are used as input of the first neural network, and the output of the first neural network is the first prediction factor.

[0010] In one embodiment of the present invention, the transmitted light intensity is input into a trained first neural network to obtain a first prediction factor, including: inputting the transmitted light intensity at time T-∆t and the continuous N-1 transmitted light intensities before this time into the trained first neural network to obtain the first prediction factor at time T; wherein, ∆t is the time for the liquid to flow from the transmitted light intensity collection point to the metal nozzle.

[0011] In one embodiment of the present invention, obtaining a second prediction factor according to the current Taylor cone angle and the target Taylor cone angle includes: obtaining the second prediction factor according to the historical data of the current Taylor cone angle and the target Taylor cone angle according to the following formula: , where B(k) is the second prediction factor, and err_θ(i) is the difference between the historical data of the target Taylor cone angle and the current Taylor cone angle.

[0012] In one embodiment of the present invention, the number of input layer nodes of the BP neural network is 4, the number of hidden layer nodes is 40, and the number of output layer nodes is 3; the target Taylor cone angle, the first prediction factor, the second prediction factor, and the control signal output by the PID controller are normalized and used as the input of the BP neural network, and the output of the BP neural network is the gain parameter K of the PID controller. p , K i , K d .

[0013] To achieve the above-mentioned purpose and other related purposes, the present invention also provides a control system for stabilizing the Taylor cone angle, comprising: a metal nozzle, for spraying liquid to form microspheres or microcapsules; a syringe pump, for extracting liquid from a liquid storage tank and delivering it to the metal nozzle; a conduit, for connecting the liquid storage tank and the syringe pump, and the syringe pump and the metal nozzle; an equipotential ring, located below the metal nozzle; a high-voltage generator, the positive and negative electrodes of the output end of the high-voltage generator are respectively connected to the metal nozzle and the equipotential ring; a collector, located below the equipotential ring, for receiving the material sprayed from the metal nozzle; a light intensity collection unit, the light intensity collection unit is arranged at the conduit at the tail of the metal nozzle, and includes a transmitting end and a receiving end, and the light emitted by the transmitting end is absorbed by the receiving end after passing through the conduit; a microscope lens, for taking an image of the Taylor cone angle below the metal nozzle; and a control unit, the control unit is connected to the syringe pump, the high-voltage generator, the light intensity collection unit and the microscope lens, and is used to execute the method as described above.

[0014] Beneficial effects of the present invention: The present invention proposes a control method and system for stabilizing the Taylor cone angle. The method obtains a first prediction factor by acquiring the transmitted light intensity of the liquid flowing through the catheter and a first neural network, and uses the first prediction factor as the input of the BP neural network, and combines the PID controller based on the BP neural network to control the output voltage of the high-voltage generator. This can stabilize the Taylor cone angle and ensure the stability of the spray solution at the metal nozzle, thereby avoiding the problem of uneven microsphere particle size or uneven microcapsule shell thickness. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1A flow chart of a Taylor cone angle control method provided by an embodiment of the present invention;

[0017] Figure 2 A flow chart for obtaining a preset flow rate and a preset voltage provided by an embodiment of the present invention;

[0018] Figure 3 A flowchart for obtaining a stable jet injection interval provided by an embodiment of the present invention;

[0019] Figure 4 A schematic diagram of a scanning point and a stable jet injection interval provided by an embodiment of the present invention;

[0020] Figure 5 A schematic diagram of a PID controller based on a BP neural network provided by an embodiment of the present invention;

[0021] Figure 6 A schematic diagram of a Taylor cone angle control system provided by an embodiment of the present invention.

[0022] Explanation of the reference numerals: 10, metal nozzle; 20, injection pump; 30, catheter; 40, equipotential ring; 50, high voltage generator; 60, collector; 70, light intensity collection unit; 80, microscope lens; 90, control unit. DETAILED DESCRIPTION

[0023] The following describes the embodiments of the present invention through specific specific embodiments, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. It should be noted that the following embodiments and the features in the embodiments can be combined with each other without conflict. In addition to the specific methods, equipment, and materials used in the embodiments, according to the technical personnel in this technical field's grasp of the prior art and the records of the present invention, any methods, equipment, and materials of the prior art that are similar or equivalent to the methods, equipment, and materials described in the embodiments of the present invention can also be used to implement the present invention.

[0024] It should be understood that the terms used in the embodiments of the present invention are for describing specific embodiments rather than for limiting the scope of protection of the present invention. Unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those generally understood by those skilled in the art.

[0025] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read, and are not used to limit the limiting conditions under which the present invention can be implemented, so they have no substantive technical significance. Any modification of the structure, change in the proportional relationship, or adjustment of the size should still fall within the scope of the technical content disclosed by the present invention without affecting the effects and purposes that can be achieved by the present invention. In addition, the drawings only show the components related to the present invention rather than the number, shape, and size of the components in actual implementation. The type, quantity, and proportion of each component in actual implementation can be changed arbitrarily, and the component layout type may also be more complicated.

[0026] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In some of the embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0027] The flowchart and block diagram in the accompanying drawings illustrate the system architecture, functions and operations that may be implemented according to the methods and computer program products of various embodiments disclosed in the present invention. In this regard, each box in the flowchart or block diagram can represent a module, a program segment, or a part of a code, and the module, the program segment, or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of the boxes in the block diagram and / or the flowchart can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0028] The Taylor cone is formed by the acceleration and contraction of liquid under an electric field. The end is very sharp, and what is ejected from the tip of the cone is a micro-nano-scale liquid jet, the diameter of which can be as small as nanometers or even atomic levels. The stability of the Taylor cone angle is the key to forming a stable jet. For a uniform and single solution, the stability of the Taylor cone angle is relatively easy to achieve; but if the solution contains particles, such as cells, silica, etc., different particle morphologies (distribution uniformity and cell state) will cause changes in the viscosity and concentration of the solution during the experiment, which will in turn affect the morphology of the Taylor cone, especially the Taylor cone angle. When the Taylor cone angle is affected, it will directly lead to uneven microsphere particle size or uneven microcapsule shell thickness. The present invention avoids these situations by stabilizing the Taylor cone angle. At the same time, the Taylor cone angle referred to in the present invention is the angle at the top of the Taylor cone, not the angle between the edge of the Taylor cone and the center line.

[0029] See also Figure 1 , Figure 1 A control method for stabilizing a Taylor cone angle provided in an embodiment of the present invention includes steps S100 to S500.

[0030] Step S100, control the particle-containing liquid to be sprayed from the metal nozzle 10 at a preset flow rate and a preset voltage. The control scheme in the present invention is performed in real time during the operation of the metal nozzle 10, so at the beginning, the device is operated at a preset flow rate and voltage.

[0031] See also Figure 2 In a specific embodiment of the present invention, the preset flow rate and the preset voltage are obtained according to steps S110 to S130.

[0032] Step S110, obtaining a plurality of scanning points according to the flow rate interval and the voltage interval. Generally speaking, the flow rate of the injection pump 20 and the output voltage of the high voltage generator 50 are limited. The flow rate interval of the injection pump 20 is [flowrate min ,flowrate max ], the output voltage range of the high voltage generator 50 is [V min ,V max ], then multiple scanning points can be obtained according to the interval range. Scanning points are virtual points, and each scanning point corresponds to a flow rate and voltage. Assuming that the voltage is the horizontal axis and the flow rate is the vertical axis, a two-dimensional plane coordinate system is constructed, then the scanning points are equivalent to multiple points evenly distributed in the above interval. Figure 4 The scanning points are evenly distributed in rows and columns. Assuming that the scanning increments of flow rate and voltage are recorded as ∆flowrate and ∆V respectively, the coordinates of each scanning point are: (flowrate min ,V min )、(flowratemin ,V min +∆V)、(flowrate min ,V min +2×∆V)、…、(flowrate min ,V max )、(flowrate min +∆flowrate,V min )、(flowrate min +∆flowrate,V min +∆V),…, this scanning method is to fix the flow rate first, then adjust the voltage, and when the voltage is scanned from the minimum value to the maximum value, increase the flow rate and scan the voltage again. In addition, you can also fix the voltage first and scan the flow rate.

[0033] Step S120, according to multiple scanning points, obtain the stable jet injection interval of the liquid without particles. Each scanning point corresponds to a flow rate and a voltage, so the stable jet injection interval of the liquid at each scanning point can be calculated. Generally, the stable jet injection interval of the liquid containing particles is inside the stable jet injection interval of the liquid without particles, so in this step, by calculating the stable jet injection interval of the liquid without particles, it is used as a reference for the preset flow rate and preset voltage of the liquid containing particles in turn, so that the Taylor cone angle can be controlled to a stable state more quickly.

[0034] See also Figure 3 and Figure 4 In a specific embodiment of the present invention, step S120 includes steps S121 to S123.

[0035] Step S121 , controlling the particle-free liquid to be sprayed from the metal nozzle 10 according to the flow rate and voltage corresponding to each scanning point.

[0036] Step S122, for any scanning point: obtain an image of the Taylor cone angle below the metal nozzle 10, and process the image to obtain the Taylor cone angle; when the change amplitude or change rate of the Taylor cone angle within the preset time is less than or equal to the preset value, the scanning point is a Class I scanning point; when the change amplitude or change rate of the Taylor cone angle within the preset time is greater than the preset value, the scanning point is a Class II scanning point. The purpose of this step is to find a scanning point located in the stable jet injection interval. In this step, the change amplitude of the Taylor cone angle corresponds to a fixed angle, and the change rate of the Taylor cone angle is related to the average angle of the Taylor cone angle. Taking the change rate of the Taylor cone angle as an example, the judgment condition for a Class I scanning point can be: the difference between the maximum or minimum value of the Taylor cone angle and the average value within 10s is less than or equal to 10% of the average value.

[0037] Step S123, obtaining a stable jet injection interval according to the first type of scanning points and the second type of scanning points, wherein the first type of scanning points are scanning points located in the stable jet injection interval, and the second type of scanning points are scanning points located outside the stable jet injection interval.

[0038] It can be understood that, since the first type of scanning points and the second type of scanning points together constitute a scanning point set, it is also possible to obtain only the first type of scanning points or only the second type of scanning points.

[0039] Step S130, according to the flow rate and voltage corresponding to the center point of the stable jet injection interval, a preset flow rate and a preset voltage are obtained. The stable jet injection interval is an area, which includes multiple scanning points. The center point of the area can be calculated. The center point does not have to be particularly accurate. It can be a scanning point located near the center position, or it can be an actual center point (not necessarily a scanning point) calculated according to the mathematical method of the area contour. Then, the flow rate and voltage corresponding to the center point are used as the preset flow rate and preset voltage.

[0040] It should be noted that the steps for determining the preset flow rate and the preset voltage utilize a liquid without particles, while step S100 is performed using a liquid containing particles.

[0041] It can be understood that the stable jet injection interval of the liquid containing particles is generally located inside the stable jet injection interval of the liquid without particles. Therefore, in this embodiment, by obtaining the center point of a type of scanning point interval as the preset flow rate and preset voltage, the system can be operated more quickly in the stable jet injection interval of the liquid containing particles, and the subsequent fine-tuning is also basically in the stable jet injection interval. In some special cases, the stable jet injection interval of the liquid containing particles may have some areas located outside the stable jet injection interval of the liquid without particles, that is, the two are in an intersecting state, and this will not affect the stability of the Taylor cone angle, because the above-mentioned preset flow rate and preset voltage are only used as initial values, which will not affect the stable control of the Taylor cone angle, but will only affect the stable adjustment time. Therefore, even in some special cases, it is feasible to select the center point of a type of scanning point interval as the preset flow rate and preset voltage.

[0042] Step S200, obtaining the transmitted light intensity of the particle-containing liquid flowing through the conduit 30, and inputting the transmitted light intensity into the trained first neural network to obtain the first prediction factor. As mentioned above, the particles contained in the liquid flowing through the conduit 30 cause the instability of the Taylor cone angle, which also causes the instability of the transmitted light intensity, and the change of the transmitted light intensity indirectly reflects the change of the Taylor cone angle. Specifically, the liquid is generally transparent. When it does not contain particles, its transmitted light intensity is basically fixed. When the liquid contains particles, the particles will cause a certain attenuation of the transmitted light, resulting in a decrease in the transmitted light intensity. The more particles there are, the greater the attenuation of the transmitted light intensity. At the same time, the viscosity and concentration of the solution also change, resulting in a change in the Taylor cone angle. Therefore, in this step, the first prediction factor is calculated by the transmitted light intensity. The first prediction factor can be, for example, a predicted Taylor cone angle, or a parameter calculated based on the Taylor cone angle.

[0043] In a specific embodiment of the present invention, the sampling frequency of the transmitted light intensity is N times the sampling frequency of the image, where N is a positive integer; the number of input layer nodes of the first neural network is N, the number of intermediate layer nodes is 60, and the number of output layer nodes is 1. N consecutive transmitted light intensities are used as the input of the first neural network, and the output of the first neural network is the first prediction factor. For example, the sampling frequency of the transmitted light intensity is fs_b, and the sampling frequency of the image is fs_a. Each image corresponds to a Taylor cone angle, so it can be understood that the sampling frequency of the Taylor cone angle is also fs_a, and the first prediction factor is calculated based on the Taylor cone angle, so the sampling frequency of the first prediction factor is also fs_a. The first neural network mainly establishes a mapping relationship between N transmitted light intensities and 1 first prediction factor, and then the value of 1 first prediction factor can be generated based on the N transmitted light intensities.

[0044] Understandably, the first neural network needs to be trained. Continuously collect N transmitted light intensities of a certain section of particle-containing liquid; input the N transmitted light intensities into the first neural network to obtain the predicted value of the first prediction factor; take a picture of the Taylor cone angle formed by the section of particle-containing liquid under the metal nozzle 10, and process it to obtain the true value of the first prediction factor; optimize the parameters of the first neural network by calculating the loss value between the predicted value and the true value of the first prediction factor, until the optimization is completed to obtain a trained first neural network.

[0045] In a specific embodiment of the present invention, the transmitted light intensity is input into the trained first neural network to obtain the first prediction factor, including: inputting the transmitted light intensity at time T-∆t and the continuous N-1 transmitted light intensities before the time into the trained first neural network to obtain the first prediction factor at time T; wherein ∆t is the time for the liquid to flow from the transmitted light intensity collection point to the bottom (tail) of the metal nozzle 10. Since the first neural network reflects the mapping relationship between the transmitted light intensity and the first prediction factor, it takes a certain amount of time, namely ∆t, for the particle-containing liquid to flow to the metal nozzle 10. ∆t can be calculated based on the distance between the transmitted light intensity collection point and the bottom (tail) of the metal nozzle 10, and the flow rate of the liquid. The introduction of ∆t can make the prediction result of the first neural network more accurate.

[0046] The setting of the above-mentioned first neural network, on the one hand, can increase the first prediction factor as the input of the subsequent BP neural network, and on the other hand, can realize the neural network input control of the multi-frequency sampling system. Since the sampling frequencies of various physical quantities in the feedback physical quantities monitored in real time are different in the same control model (for example, the sampling frequencies of the transmitted light intensity and the image are different), the input neural network cannot perform operations on different time scales during the operation (that is, multiple transmitted light intensity data cannot be input into the BP neural network). Consider adding a variable frequency neural network (that is, the first neural network, whose number of input quantities N is adjustable) for operation, and output data synchronized with other parameters into the BP neural network to realize variable frequency processing of data throughput.

[0047] Step S300, obtain an image of the Taylor cone angle below the metal nozzle 10, and process the image to obtain the current Taylor cone angle. For example, the image can be binarized, and then the two sides of the Taylor cone can be obtained by linear fitting, and then the angle between the two sides can be solved to obtain the Taylor cone angle. This image processing process is very common and will not be described in detail.

[0048] Before explaining steps S400 and S500, a PID controller based on BP neural network is briefly introduced.

[0049] PID controller is a common feedback loop component in industrial control applications. This controller compares the collected data with a reference value, and then uses the difference to calculate a new input value. The purpose of this new input value is to make the system data reach or remain at the reference value. The PID controller consists of a proportional unit P, an integral unit I and a differential unit D. p , K i and K d Three parameters are set to realize the control of the system.

[0050] In the early PID controller, Kp , K i and K d The three parameters are fixed values, and their control effect on the dynamic transfer function system is not ideal. Subsequently, a PID controller based on the BP neural network was proposed. The BP neural network can automatically adjust the three parameters of the PID controller according to the real-time feedback of the system, so that the control system can adapt to changes in the environment and load, and improve the flexibility and accuracy of the control. Once the parameters of the traditional PID controller are set, manual intervention is usually required to adjust them, which may not be timely or accurate enough in a rapidly changing industrial environment. In addition, the BP neural network can capture and approximate complex nonlinear relationships through learning, which is particularly effective for processing nonlinear systems. In contrast, the traditional PID controller is based on linear control theory, and it may be difficult to achieve ideal control effects for nonlinear systems.

[0051] In the common PID controller structure based on BP neural network, the input of BP neural network generally includes expected value, error signal, control amount, etc. Among them, the expected value is also called the target value, which refers to the expected stable value of Taylor cone angle in the present invention, denoted as r_θ(k); the error signal is calculated based on the output of the system and the expected value, wherein the output of the system is the actual Taylor cone angle obtained according to the image in the present invention, denoted as y_θ(k), and the error signal err_θ(k)=r_θ(k)-y_θ(k); the control amount is the control signal of the high-voltage generator 50 to be controlled in the present invention, denoted as u_θ(k). The first prediction factor is introduced in the present invention, so the input of BP neural network is also adjusted, which will be described in detail below. The PID controller based on BP neural network has a continuous learning and adjustment process, so it involves time series, wherein r_θ(k) generally does not change with the change of k, while y_θ(k) and u_θ(k) change with the increase of k.

[0052] Step S400: Obtain a second prediction factor according to the current Taylor cone angle and the target Taylor cone angle. The second prediction factor may be, for example, the error signal err_θ(k) mentioned above.

[0053] In a specific embodiment of the present invention, step S400 includes: obtaining a second prediction factor according to the historical data of the current Taylor cone angle and the target Taylor cone angle according to the following formula:

[0054] , where B(k) is the second prediction factor, and err_θ(i) is the difference between the historical data of the target Taylor cone angle and the current Taylor cone angle, that is, err_θ(i)=r_θ(i)-y_θ(i). In this embodiment, the historical data of the Taylor cone angle is introduced to calculate the second prediction factor, and the prediction function of the time series correlation of the three parameters of the PID controller is added to further improve the control effect.

[0055] See also Figure 5 , step S500, using a PID controller based on a BP neural network to control the output voltage of the high voltage generator 50 to stabilize the Taylor cone angle; wherein the input of the BP neural network includes the target Taylor cone angle, the first prediction factor, the second prediction factor, and the control signal output by the PID controller. The difference between this step and other control processes of the PID controller based on the BP neural network is that the input of the BP neural network is different, and based on these inputs, the control of the Taylor cone angle will be more convenient.

[0056] It should be noted that the output voltage of the high-voltage generator 50 is generally thousands of volts, and the control signal output by the PID controller is, for example, a value of 0 to 1 (or 0 to 10). The control signal output by the PID controller is not equal to the output voltage of the high-voltage generator 50. The output of the PID controller is multiplied by a constant to obtain the output voltage of the high-voltage generator 50.

[0057] Figure 5 In the figure, NN_A is the first neural network, NN is the BP neural network, PID is the corresponding PID controller, and the high voltage source is the high voltage generator 50.

[0058] In a specific embodiment of the present invention, the number of input layer nodes of the BP neural network is 4, the number of hidden layer nodes is 40, and the number of output layer nodes is 3; the target Taylor cone angle r_θ(k), the first prediction factor A(k), the second prediction factor B(k) and the voltage u_θ(k) output by the PID controller are normalized and used as the input of the BP neural network, and the output of the BP neural network is the gain parameter K of the PID controller. p , K i , K d Through the self-learning ability of BP neural network, the three parameters K p , K i , K d The PID controller is adjusted to an optimal state, wherein u_θ(k) is a control signal output by the PID controller, and is used to control the output voltage of the high voltage generator 50 .

[0059] See also Figure 6 , Figure 6A control system for stabilizing the Taylor cone angle provided in one embodiment of the present invention includes a metal nozzle 10, an injection pump 20, a catheter 30, an equipotential ring 40, a high voltage generator 50, a collector 60, a light intensity collection unit 70, a microscope lens 80 and a control unit 90. Among them, the metal nozzle 10 is used to spray liquid to form microspheres or microcapsules; the injection pump 20 is used to extract liquid from the liquid storage tank and transport it to the metal nozzle 10; the catheter 30 is used to connect the liquid storage tank and the injection pump 20, and the injection pump 20 and the metal nozzle 10; the equipotential ring 40 is located below the metal nozzle 10; the positive and negative poles of the output ends of the high-voltage generator 50 are respectively connected to the metal nozzle 10 and the equipotential ring 40; the collector 60 is located below the equipotential ring 40 and is used to receive the material sprayed from the metal nozzle; the light intensity collection unit 70 is arranged at the catheter 30 at the tail of the metal nozzle 10, which includes a transmitting end and a receiving end. The light emitted by the transmitting end is absorbed by the receiving end after passing through the catheter 30. The transmitting end is mainly used to emit light, and the receiving end is used to receive the light transmitted from the catheter 30, and the light received is collected by the receiving end. The light signal is converted into a voltage signal. The magnitude of the voltage directly reflects the intensity of the light. The transmitted light intensity can be directly analyzed according to the voltage signal. The transmitting end can be, for example, an infrared LED light emitting diode, and the receiving end can be, for example, an infrared photodiode; the microscope lens 80 is used to take an image of the Taylor cone angle under the metal nozzle; the control unit 90 is connected to the injection pump 20 for controlling the liquid flow rate, the control unit 90 is connected to the high-voltage generator 50 for controlling the voltage, the control unit 90 is connected to the light intensity acquisition unit 70 for acquiring light intensity data, and the control unit 90 is also connected to the microscope lens 80 for acquiring the image of the Taylor cone angle. The control unit 90 also controls the voltage of the high-voltage generator 50 according to the acquired data according to the method mentioned above to stabilize the Taylor cone angle. The specific control method has been described in detail above and will not be repeated here.

[0060] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A control method for stabilizing the Taylor cone angle, characterized in that: include: Controlling the particle-containing liquid to be ejected from the metal nozzle according to a preset flow rate and a preset voltage; Acquiring the transmitted light intensity of the particle-containing liquid flowing through the conduit, and inputting the transmitted light intensity into a trained first neural network to obtain a first prediction factor; Acquire an image of the Taylor cone angle below the metal nozzle, and process the image to obtain a current Taylor cone angle; Obtaining a second prediction factor according to the current Taylor cone angle and the target Taylor cone angle; The output voltage of the high-voltage generator is controlled by a PID controller based on a BP neural network to stabilize the Taylor cone angle; wherein the input of the BP neural network includes the target Taylor cone angle, the first prediction factor, the second prediction factor and the control signal output by the PID controller.

2. The control method for stabilizing the Taylor cone angle according to claim 1, characterized in that: The preset flow rate and the preset voltage are obtained according to the following steps: According to the flow rate interval and the voltage interval, multiple scanning points are obtained; According to the plurality of scanning points, a stable jet injection interval of the liquid without particles is obtained; The preset flow rate and the preset voltage are obtained according to the flow rate and the voltage corresponding to the center point of the stable jet injection interval.

3. The control method for stabilizing the Taylor cone angle according to claim 2, characterized in that: According to the plurality of scanning points, a stable jet injection interval of the liquid without particles is obtained, including: Controlling the particle-free liquid to be ejected from the metal nozzle according to the flow rate and voltage corresponding to each scanning point; For any scan point: Acquire an image of the Taylor cone angle below the metal nozzle, and process the image to obtain the Taylor cone angle; When the change amplitude or change rate of the Taylor cone angle within the preset time is less than or equal to the preset value, the scanning point is a type I scanning point; When the change amplitude or change rate of the Taylor cone angle within the preset time is greater than the preset value, the scanning point is a Class II scanning point; The stable jet injection interval is obtained according to the first type of scanning points and the second type of scanning points.

4. The control method for stabilizing the Taylor cone angle according to claim 1, characterized in that: The sampling frequency of the transmitted light intensity is N times the sampling frequency of the image, where N is a positive integer; the number of input layer nodes of the first neural network is N, the number of intermediate layer nodes is 60, and the number of output layer nodes is 1, and N consecutive transmitted light intensities are used as the input of the first neural network, and the output of the first neural network is the first prediction factor.

5. The control method for stabilizing the Taylor cone angle according to claim 4, characterized in that: The transmitted light intensity is input into the trained first neural network to obtain a first prediction factor, including: The transmitted light intensity at time T-∆t and the continuous N-1 transmitted light intensities before this time are input into the trained first neural network to obtain the first prediction factor at time T; wherein, ∆t is the time for the liquid to flow from the transmitted light intensity collection point to the metal nozzle.

6. The control method for stabilizing the Taylor cone angle according to claim 1, characterized in that: According to the current Taylor cone angle and the target Taylor cone angle, a second prediction factor is obtained, including: According to the historical data of the current Taylor cone angle and the target Taylor cone angle, the second prediction factor is obtained according to the following formula: , where B(k) is the second prediction factor, and err_θ(i) is the difference between the historical data of the target Taylor cone angle and the current Taylor cone angle.

7. The control method for stabilizing the Taylor cone angle according to claim 1, characterized in that: The number of input layer nodes of the BP neural network is 4, the number of hidden layer nodes is 40, and the number of output layer nodes is 3; the target Taylor cone angle, the first prediction factor, the second prediction factor, and the control signal output by the PID controller are normalized and used as the input of the BP neural network, and the output of the BP neural network is the gain parameter K of the PID controller. p , K i , K d .

8. A control system for stabilizing the Taylor cone angle of an electrofluid jet, characterized in that: include: Metal nozzles for spraying liquid to form microspheres or microcapsules; A syringe pump is used to extract liquid from the liquid storage tank and transport it to the metal nozzle; A conduit, used for connecting the liquid storage tank and the injection pump, and the injection pump and the metal nozzle; An equipotential ring, located below the metal nozzle; A high voltage generator, wherein the positive and negative electrodes of the output terminals of the high voltage generator are respectively connected to the metal nozzle and the equipotential ring; A collector, located below the equipotential ring, for receiving the material ejected from the metal nozzle; A light intensity collection unit, which is arranged at the conduit at the tail of the metal nozzle, and comprises a transmitting end and a receiving end, and the light emitted by the transmitting end passes through the conduit and is absorbed by the receiving end; A microscope lens, used for taking an image of the Taylor cone angle below the metal nozzle; and a control unit, wherein the control unit is connected to the injection pump, the high voltage generator, the light intensity collection unit and the microscope lens, and is used to execute the method according to any one of claims 1 to 7.

Images