CN113285595A - PID parameter setting system and control method of digital power supply based on machine learning - Google Patents

PID parameter setting system and control method of digital power supply based on machine learning Download PDF

Info

Publication number
CN113285595A
CN113285595A CN202110642069.4A CN202110642069A CN113285595A CN 113285595 A CN113285595 A CN 113285595A CN 202110642069 A CN202110642069 A CN 202110642069A CN 113285595 A CN113285595 A CN 113285595A
Authority
CN
China
Prior art keywords
machine learning
digital
pid
module
power supply
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110642069.4A
Other languages
Chinese (zh)
Inventor
不公告发明人
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhuhai Amicro Semiconductor Co Ltd
Original Assignee
Zhuhai Amicro Semiconductor Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhuhai Amicro Semiconductor Co Ltd filed Critical Zhuhai Amicro Semiconductor Co Ltd
Priority to CN202110642069.4A priority Critical patent/CN113285595A/en
Publication of CN113285595A publication Critical patent/CN113285595A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
    • H02M3/00—Conversion of dc power input into dc power output
    • H02M3/02—Conversion of dc power input into dc power output without intermediate conversion into ac
    • H02M3/04—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters
    • H02M3/10—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode
    • H02M3/145—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal
    • H02M3/155—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only
    • H02M3/156—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators
    • H02M3/157—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators with digital control
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B11/00—Automatic controllers
    • G05B11/01—Automatic controllers electric
    • G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
    • G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P. I., P. I. D.
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
    • H02M1/00—Details of apparatus for conversion
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
    • H02M3/00—Conversion of dc power input into dc power output
    • H02M3/02—Conversion of dc power input into dc power output without intermediate conversion into ac
    • H02M3/04—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters
    • H02M3/10—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode
    • H02M3/145—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal
    • H02M3/155—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only
    • H02M3/156—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators
    • H02M3/158—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators including plural semiconductor devices as final control devices for a single load
    • H02M3/1582—Buck-boost converters

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Dc-Dc Converters (AREA)

Abstract

The invention discloses a PID parameter setting system and a control method of a digital power supply based on machine learning, wherein the system comprises a machine learning module, a digital controller module and a power circuit module; the machine learning unit is used for adjusting the PID parameters of the digital controller module according to the voltage difference value change information and the adjustment record sent by the digital controller module; the load disturbance module is used for generating a disturbance signal to change the voltage output by the power circuit module; the digital controller module is used for outputting a corresponding pulse signal according to a voltage difference value between the voltage output by the power circuit module and the reference voltage of the digital controller module and the adjusted PID parameter; the power circuit module outputs corresponding voltage according to the pulse signal sent by the digital controller module, and the system can quickly and accurately set PID parameters according to different application requirements, so that the digital power system always operates in the optimal state.

Description

PID parameter setting system and control method of digital power supply based on machine learning
Technical Field
The invention relates to the technical field of digital power supply controllers, in particular to a PID parameter setting system and a control method of a digital power supply based on machine learning.
Background
The digital power supply refers to the external characteristics of a switching power supply, and functions such as controlling, managing, monitoring the power supply and the like can be realized by using a Digital Signal Processor (DSP) or a Micro Controller Unit (MCU) and using a digital power supply driver as a control object. The digital power supply has the advantages of high integration, small volume, strong anti-interference capability, strong stability and reliability, flexible control mode, easy intellectualization and the like, and is an important development direction in the field of power supplies. The digital controller is a core module of the digital power supply and is responsible for generating a control signal of the switching power supply, and the performance of the digital controller directly influences the stability and the dynamic response speed of the whole power supply, so that the design of the high-performance digital controller has important significance.
In the loop control of the digital power supply, a PID (proportional-integral-differential) control algorithm is generally adopted, and the PID control algorithm is the most important parameter setting in practical application. In general, the setting of parameters is highly dependent on the experience of engineering technicians, and the actually applied digital power supply systems are very different, and factors such as nonlinearity, time variation, large hysteresis and the like exist. The traditional setting method is time-consuming and labor-consuming, and the PID parameter obtained by setting is often not an ideal value, so that the power supply system is naturally difficult to stably work near a set value.
Disclosure of Invention
In order to solve the problems, the invention discloses a PID parameter setting system and a control method of a digital power supply based on machine learning, and the method can quickly and accurately set PID parameters so that the digital power supply system can operate in an optimal state. The specific technical scheme is as follows:
a PID parameter setting system of a digital power supply based on machine learning comprises a machine learning module, a digital controller module and a power circuit module, wherein the machine learning module comprises a machine learning unit and a load disturbance unit, the machine learning unit is respectively connected with the load disturbance unit and the digital controller module, the load disturbance unit is respectively connected with the machine learning unit and the power circuit module, and the digital controller module is respectively connected with the machine learning unit and the power circuit module; the machine learning unit is used for adjusting the PID parameters of the digital controller module according to the voltage difference value change information and the adjustment record sent by the digital controller module; the load disturbance module is used for generating a disturbance signal to change the voltage output by the power circuit module; the digital controller module is used for outputting a corresponding pulse signal according to a voltage difference value between the voltage output by the power circuit module and the reference voltage of the digital controller module and the adjusted PID parameter; and the power circuit module outputs corresponding voltage according to the pulse signal sent by the digital controller module.
Compared with the prior art, in the parameter learning process of the machine learning system, the load disturbance unit continuously generates disturbance signals to change the load, so that the output voltage of the power supply is unstable, and then PID control parameters are repeatedly configured through the machine learning device to quickly recover the output voltage to a stable state; the load disturbance unit well simulates unstable factors of the power supply in practical application, and the machine learning system actively generates disturbance signals to repeatedly learn PID parameters controlled by the power supply, so that the optimal control parameters can be quickly found.
Further, the digital controller module comprises an analog-to-digital converter, a DPID controller and a digital pulse broadband modulator which are connected in sequence, the analog-to-digital converter is used for converting the voltage difference signal into a discrete difference signal, the DPID controller is used for generating a discrete control signal according to the PID parameter and the discrete difference signal, and the digital pulse broadband modulator is used for converting the discrete control signal into a pulse signal with a corresponding duty ratio.
Furthermore, a variable resistor is arranged in the power circuit module, the load disturbance unit is a signal sending circuit, and the signal sending circuit is used for sending a disturbance signal to the variable resistor to change the resistance value of the variable resistor, so that the voltage output by the power circuit module changes.
Further, the machine learning unit is an intelligent device having the capability of detecting, comparing and recording voltage difference value change information and setting and adjusting PID parameters.
A control method of a PID parameter setting system of a digital power supply based on machine learning is used for controlling the PID parameter setting system of the digital power supply based on machine learning, and comprises the following steps: s1: configuring an initial value and a boundary condition of a PID parameter by a machine learning unit; s2: the machine learning unit enables the load disturbance module to send a disturbance signal to the power circuit module, so that the voltage output by the power circuit module fluctuates; s3: the machine learning unit acquires voltage difference value change information between the voltage output by the power circuit module and a reference voltage of the digital controller module in the fluctuation process; s4: the machine learning unit records current voltage difference value change information and corresponding PID parameters, and judges whether the current PID parameters have the optimal control effect or not according to the current power supply difference value change information; s5: if not, adjusting the PID parameters according to the adjustment history of the PID parameters and the recorded voltage difference value change information, and if so, setting the current PID parameters as the optimal PID parameters and storing the optimal PID parameters; s6: and repeating the steps S2 to S5 until the optimal PID parameters are found.
Compared with the prior art, in the parameter learning process of the machine learning system, the load disturbance unit continuously generates disturbance signals to make the output voltage of the power supply unstable, and then the PID control parameters are repeatedly configured through the machine learning device to make the output voltage quickly recover to a stable state; the load disturbance unit well simulates unstable factors of the power supply in practical application, and actively generates disturbance signals through a machine learning system, so that PID parameters controlled by the power supply are repeatedly learned, and the optimal control parameters are quickly found. Machine learning is continuously carried out through equipment, PID parameters of the digital power supply are automatically adjusted, one set of PID control parameters can be respectively adjusted under different digital power supplies or different working settings, the optimal PID parameters are automatically matched under different power supply requirements, the operation is simple, manual debugging of the PID parameters of the digital power supply is not needed, the time and the energy of engineering technicians are saved, machine learning is continuously carried out through the equipment, the parameters are closer to the optimal solution of a control system, and the accuracy of the PID parameters is guaranteed.
Further, the PID parameters include a proportional parameter, an integral parameter, and a differential parameter.
Further, in step S2, the disturbance signal generated by the load disturbance unit fluctuates the voltage output by the power circuit module by changing the resistance value of the resistor in the power circuit module. The output voltage of the power circuit module is influenced by changing the resistance value of the resistor, and the power circuit module is simple in structure and easy to realize.
Further, in step S3, after receiving the voltage output by the power circuit module, the digital controller module compares the voltage with its own reference voltage to generate a voltage difference signal, and then converts the voltage difference signal into a discrete difference signal through the analog-to-digital converter to send to the machine learning unit and the DPID controller, and the DPID controller calculates a discrete control signal according to the current PID parameter and the discrete difference signal and sends the discrete control signal to the digital pulse wideband modulator, and the digital pulse wideband modulator converts the discrete control signal into a pulse signal with a corresponding duty ratio and sends the pulse signal to the power circuit module, so that the voltage output by the power circuit module is stabilized within a preset range of the reference voltage.
Further, in step S3, the voltage difference variation information is the fluctuation amplitude and the fluctuation time of the discrete difference signal received by the machine learning unit.
Further, in step S5, if the maximum fluctuation amplitude and the fluctuation time are both less than or equal to the set values, it is determined that the current PID parameter has the optimal control effect.
Further, in step S5, if there is no history of PID parameter adjustment, the PID parameters are randomly adjusted within the boundary conditions.
Drawings
FIG. 1 is a schematic structural diagram of a PID parameter tuning system of a digital power supply based on machine learning according to an embodiment of the invention;
FIG. 2 is a flow chart of a control method for a PID parameter tuning system of a digital power supply based on machine learning according to an embodiment of the invention;
fig. 3 is a fluctuation diagram of the power supply output voltage of the PID parameter tuning system of the digital power supply based on machine learning according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the present application. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments provided in the present application without any inventive step are within the scope of protection of the present application.
Referring to fig. 1, a PID parameter tuning system of a digital power supply based on machine learning includes a machine learning module, a digital controller module and a power circuit module, where the machine learning module includes a machine learning unit and a load disturbance unit, the machine learning unit is connected to the load disturbance unit and the digital controller module, the load disturbance unit is connected to the machine learning unit and the power circuit module, and the digital controller module is connected to the machine learning unit and the power circuit module; the machine learning unit is used for adjusting the PID parameters of the digital controller module according to the voltage difference value change information and the adjustment record sent by the digital controller module; the load disturbance module is used for generating a disturbance signal to change the voltage output by the power circuit module; the digital controller module is used for outputting a corresponding pulse signal according to a voltage difference value between the voltage output by the power circuit module and the reference voltage of the digital controller module and the adjusted PID parameter; and the power circuit module outputs corresponding voltage according to the pulse signal sent by the digital controller module. Compared with the prior art, in the parameter learning process of the machine learning system, the load disturbance unit continuously generates disturbance signals to change the load, so that the output voltage of the power supply is unstable, and then PID control parameters are repeatedly configured through the machine learning device to quickly recover the output voltage to a stable state; the load disturbance unit well simulates unstable factors of the power supply in practical application, and the machine learning system actively generates disturbance signals to repeatedly learn PID parameters controlled by the power supply, so that the optimal control parameters can be quickly found.
As one example, the machine learning unit is an intelligent device having the capability of detecting, comparing and recording voltage difference value change information and setting and adjusting PID parameters. The power supply circuit module is internally provided with a variable resistor, the load disturbance unit is a signal sending circuit, and the signal sending circuit is used for sending a disturbance signal to the variable resistor to change the resistance value of the variable resistor, so that the voltage output by the power supply circuit module is changed. The intelligent device can be a computer, or a processor with corresponding functions, a carrier of programs or software, and the like; the variable resistor is a common variable resistor which changes a resistance value according to a signal, the load disturbance unit is a signal transmission circuit or a signal transmitter which can receive a signal of the intelligent device and then transmit a corresponding disturbance signal to change the resistance value of the variable resistor, and the connection between the variable resistor and the signal transmission circuit or the signal transmitter is a common technology and is not described herein.
As an embodiment, the digital controller module includes an analog-to-digital converter, a DPID controller, and a digital pulse wideband modulator, which are connected in sequence, where the analog-to-digital converter is configured to convert the voltage difference signal into a discrete difference signal, the DPID controller is configured to generate a discrete control signal according to the PID parameter and the discrete difference signal, and the digital pulse wideband modulator is configured to convert the discrete control signal into a pulse signal with a corresponding duty ratio. The power circuit module is a voltage boosting circuit and a voltage reducing circuit. As shown in fig. 1, the power supply circuit module is a power supply BUCK circuit module, that is, a step-up and step-down circuit, the power supply circuit composed of electronic components in the figure is an embodiment in the present application, the type and number of the electronic components composing the power supply circuit module are not limited, and the type, number, and resistance of the electronic components may be designed according to actual conditions as long as the step-up and step-down power supply circuit is started. Vout is the output voltage of the power circuit module, Vref is the reference voltage of the digital controller module (i.e. the set value of the output voltage), ADC is the mode converter, DPID is the DPID controller, DPWM is the digital pulse wideband modulator, e (t) is the voltage difference signal of the output voltage (i.e. the difference between the reference voltage Vref and the output voltage Vout), e (n) is the discrete value of the voltage difference signal, i.e. the discrete difference signal, u (n) is the discrete control signal output by the DPID controller. The analog-to-digital converter ADC converts the voltage difference signal e (t) into a discrete difference signal e (n), then the DPID controller calculates a corresponding discrete control signal u (n) according to the magnitude of the discrete difference signal, and the digital pulse wideband modulator DPWM generates a pulse signal with a corresponding duty ratio according to the discrete control signal u (n), so that the output voltage of Vout is stabilized near the Vref reference voltage.
As shown in fig. 2, a method for controlling the PID parameter tuning system of the machine learning-based digital power supply, the method is used for controlling the PID parameter tuning system of the machine learning-based digital power supply, and the method comprises the following steps: s1: configuring an initial value and a boundary condition of a PID parameter by a machine learning unit; s2: the machine learning unit enables the load disturbance module to send a disturbance signal to the power circuit module, so that the voltage output by the power circuit module fluctuates; s3: the machine learning unit acquires voltage difference value change information between the voltage output by the power circuit module and a reference voltage of the digital controller module in the fluctuation process; s4: the machine learning unit records current voltage difference value change information and corresponding PID parameters, and judges whether the current PID parameters have the optimal control effect or not according to the current power supply difference value change information; s5: if not, adjusting the PID parameters according to the adjustment history of the PID parameters and the recorded voltage difference value change information, and if so, setting the current PID parameters as the optimal PID parameters and storing the optimal PID parameters; s6: and repeating the steps S2 to S5 until the optimal PID parameters are found. Compared with the prior art, in the parameter learning process of the machine learning system, the load disturbance unit continuously generates disturbance signals to make the output voltage of the power supply unstable, and then the PID control parameters are repeatedly configured through the machine learning device to make the output voltage quickly recover to a stable state; the load disturbance unit well simulates unstable factors of the power supply in practical application, and actively generates disturbance signals through a machine learning system, so that PID parameters controlled by the power supply are repeatedly learned, and the optimal control parameters are quickly found. Machine learning is continuously carried out through equipment, PID parameters of the digital power supply are automatically adjusted, one set of PID control parameters can be respectively adjusted under different digital power supplies or different working settings, the optimal PID parameters are automatically matched under different power supply requirements, the operation is simple, manual debugging of the PID parameters of the digital power supply is not needed, the time and the energy of engineering technicians are saved, machine learning is continuously carried out through the equipment, the parameters are closer to the optimal solution of a control system, and the accuracy of the PID parameters is guaranteed.
As one example, the PID parameters include a proportional parameter, an integral parameter, and a derivative parameter. In step S2, the disturbance signal generated by the load disturbance unit changes the resistance value of the resistor in the power circuit module to fluctuate the voltage output by the power circuit module. The output voltage of the power circuit module is influenced by changing the resistance value of the resistor, and the power circuit module is simple in structure and easy to realize.
As an embodiment, in step S3, after receiving the voltage output by the power circuit module, the digital controller module compares the voltage with its own reference voltage to generate a voltage difference signal, and then converts the voltage difference signal into a discrete difference signal through the analog-to-digital converter and sends the discrete difference signal to the machine learning unit and the DPID controller, and the DPID controller calculates a discrete control signal according to the current PID parameter and the discrete difference signal and sends the discrete control signal to the digital pulse wideband modulator, and the digital pulse wideband modulator converts the discrete control signal into a pulse signal with a corresponding duty ratio and sends the pulse signal to the power circuit module, so that the voltage output by the power circuit module is stabilized within a preset range of the reference voltage. In step S3, the voltage difference variation information is the fluctuation amplitude and the fluctuation time of the discrete difference signal received by the machine learning unit. In step S5, if the maximum fluctuation amplitude and the fluctuation time are both less than or equal to the set values, it is determined that the current PID parameter has the optimal control effect. In step S5, if there is no history of PID parameter adjustment, the PID parameters are adjusted within the boundary conditions at random.
As shown in fig. 3, the machine learning unit first configures the initial values and boundary conditions (i.e. the applicable ranges of the parameters) of the parameters of the PID, then the output voltage of the power circuit module fluctuates through the disturbance signal sent by the load disturbance module, the voltage difference signal e (t) obtained by the digital controller module also fluctuates, when the output voltage fluctuates, the digital controller module converts the voltage difference signal e (t) into a discrete difference signal e (n) through the analog-to-digital converter ADC, then the DPID controller calculates corresponding discrete control signals u (n) according to the PID parameters and the discrete difference signal, the digital pulse wide band modulator DPWM generates pulse signals with corresponding duty ratios according to the discrete control signals u (n), the Vout output voltage is stabilized around the Vref reference voltage and no fluctuation of the output voltage occurs. The voltage difference signal e (T) is the maximum amplitude of each oscillation, T1 is the duration of each oscillation, the machine learning unit can detect e (T) and T1 in real time, the magnitudes of e (T) and T1 are used as the judgment basis of the control effect of the current PID parameter, the smaller the numerical values of e (T) and T1 are, the more superior the corresponding PID parameter is, and the better the control effect of the PID parameter is. The machine learning unit judges whether the values of e (T) and T1 are in a set range, if the values of e (T) and T1 are not in the set range, Kp, Ki and Kd of PID parameters are respectively adjusted, during adjustment, the PID control signal can be adjusted according to the variation trends of e (T) and T1, and the disturbance output voltage of the disturbance signal is sent again to adjust the PID control signal; if the values of e (T) and T1 are in the set range, the ideal control effect is achieved, the PID parameters under the best control effect are obtained and stored, a plurality of PID parameters meeting the requirements can be obtained, and then the PID parameters with the best control effect are selected from the PID parameters meeting the requirements as the finally obtained PID parameters. By continuously generating load disturbance, analyzing the change trend of the load disturbance, reconfiguring PID parameters, recording the control effect of each parameter, and gradually improving the control effect in a better direction, the optimal parameters of the PID control system are found.
The features of the above embodiments may be arbitrarily combined, and for the sake of brevity, all possible combinations of the above embodiments are not described, but should be considered as within the scope of the present specification as long as there is no contradiction between the combinations of the features.
The above embodiments only express a few embodiments of the present invention, and the description thereof is specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application.

Claims (11)

1. A PID parameter setting system of a digital power supply based on machine learning is characterized by comprising a machine learning module, a digital controller module and a power circuit module, wherein the machine learning module comprises a machine learning unit and a load disturbance unit, the machine learning unit is respectively connected with the load disturbance unit and the digital controller module, the load disturbance unit is respectively connected with the machine learning unit and the power circuit module, and the digital controller module is respectively connected with the machine learning unit and the power circuit module;
the machine learning unit is used for adjusting the PID parameters of the digital controller module according to the voltage difference value change information and the adjustment record sent by the digital controller module;
the load disturbance module is used for generating a disturbance signal to change the voltage output by the power circuit module;
the digital controller module is used for outputting a corresponding pulse signal according to a voltage difference value between the voltage output by the power circuit module and the reference voltage of the digital controller module and the adjusted PID parameter;
and the power circuit module outputs corresponding voltage according to the pulse signal sent by the digital controller module.
2. The PID parameter tuning system of a machine learning-based digital power supply of claim 1, wherein the digital controller module comprises an analog-to-digital converter, a DPID controller and a digital pulse wideband modulator connected in sequence, the analog-to-digital converter is configured to convert the voltage difference signal into a discrete difference signal, the DPID controller is configured to generate a discrete control signal according to the PID parameter and the discrete difference signal, and the digital pulse wideband modulator is configured to convert the discrete control signal into a pulse signal with a corresponding duty ratio.
3. The PID parameter tuning system of a digital power supply based on machine learning of claim 1, wherein a variable resistor is disposed in the power supply circuit module, the load disturbance unit is a signal transmission circuit, and the signal transmission circuit is configured to transmit a disturbance signal to the variable resistor, so that a resistance value of the variable resistor changes, and thus a voltage output by the power supply circuit module changes.
4. The PID parameter tuning system of a machine learning-based digital power supply according to claim 1, wherein the machine learning unit is an intelligent device having the capability of detecting, comparing and recording the voltage difference value change information and setting and adjusting PID parameters.
5. A method for controlling the PID parameter tuning system of a machine learning based digital power supply according to any one of claims 1 to 4, wherein the method comprises the following steps:
s1: configuring an initial value and a boundary condition of a PID parameter by a machine learning unit;
s2: the machine learning unit enables the load disturbance module to send a disturbance signal to the power circuit module, so that the voltage output by the power circuit module fluctuates;
s3: the machine learning unit acquires voltage difference value change information between the voltage output by the power circuit module and a reference voltage of the digital controller module in the fluctuation process;
s4: the machine learning unit records current voltage difference value change information and corresponding PID parameters, and judges whether the current PID parameters have the optimal control effect or not according to the current power supply difference value change information;
s5: if not, adjusting the PID parameters according to the adjustment history of the PID parameters and the recorded voltage difference value change information, and if so, setting the current PID parameters as the optimal PID parameters and storing the optimal PID parameters;
s6: and repeating the steps S2 to S5 until the optimal PID parameters are found.
6. The method of controlling a PID parameter tuning system of a machine learning based digital power supply of claim 5, wherein the PID parameters include a proportional parameter, an integral parameter and a derivative parameter.
7. The method for controlling the PID parameter tuning system of the machine learning-based digital power supply according to claim 5, wherein in step S2, the disturbance signal sent by the load disturbance unit fluctuates the voltage output by the power circuit module by changing the resistance value of the resistor in the power circuit module.
8. The method as claimed in claim 7, wherein in step S3, the digital controller module generates a voltage difference signal by comparing with its own reference voltage after receiving the voltage output by the power circuit module, and then converts the voltage difference signal into a discrete difference signal through the analog-to-digital converter and sends the discrete difference signal to the machine learning unit and the DPID controller, and the DPID controller calculates the discrete control signal according to the current PID parameter and the discrete difference signal and sends the discrete control signal to the digital pulse wideband modulator, which converts the discrete control signal into a pulse signal with a corresponding duty ratio and sends the pulse signal to the power circuit module to stabilize the voltage output by the power circuit module within a preset range of the reference voltage.
9. The method for controlling the PID parameter tuning system of machine learning-based digital power supply according to claim 7, wherein in step S3, the voltage difference variation information is the fluctuation amplitude and the fluctuation time of the discrete difference signal received by the machine learning unit.
10. The method for controlling a PID parameter tuning system of a digital power supply based on machine learning of claim 9, wherein in step S5, if the maximum fluctuation range and the fluctuation time are both less than or equal to the set values, it is determined that the current PID parameter has the optimal control effect.
11. The method for controlling a PID parameter tuning system of a digital power supply based on machine learning according to claim 6, wherein in step S5, if there is no history of PID parameter tuning, the PID parameter is randomly tuned within the boundary condition.
CN202110642069.4A 2021-06-09 2021-06-09 PID parameter setting system and control method of digital power supply based on machine learning Pending CN113285595A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110642069.4A CN113285595A (en) 2021-06-09 2021-06-09 PID parameter setting system and control method of digital power supply based on machine learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110642069.4A CN113285595A (en) 2021-06-09 2021-06-09 PID parameter setting system and control method of digital power supply based on machine learning

Publications (1)

Publication Number Publication Date
CN113285595A true CN113285595A (en) 2021-08-20

Family

ID=77283855

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110642069.4A Pending CN113285595A (en) 2021-06-09 2021-06-09 PID parameter setting system and control method of digital power supply based on machine learning

Country Status (1)

Country Link
CN (1) CN113285595A (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008234326A (en) * 2007-03-20 2008-10-02 Masami Saeki Method and device for adjusting control parameter of pid controller
CN104638899A (en) * 2015-02-05 2015-05-20 东南大学 Quick start digital power based on integral separation structure
CN104779798A (en) * 2015-04-27 2015-07-15 东南大学 Method for controlling fuzzy PID digital control DC-DC converter
CN107783423A (en) * 2017-10-25 2018-03-09 珠海格力电器股份有限公司 PID parameter self-tuning method and device based on machine learning
CN109254528A (en) * 2018-11-29 2019-01-22 曾喆昭 The fast wisdom PID control method of one kind three
US20190187631A1 (en) * 2017-12-15 2019-06-20 Exxonmobil Research And Engineering Company Adaptive pid controller tuning via deep reinforcement learning
CN111582493A (en) * 2020-04-15 2020-08-25 马鞍山师范高等专科学校 Management system for deep learning model training

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2008234326A (en) * 2007-03-20 2008-10-02 Masami Saeki Method and device for adjusting control parameter of pid controller
CN104638899A (en) * 2015-02-05 2015-05-20 东南大学 Quick start digital power based on integral separation structure
CN104779798A (en) * 2015-04-27 2015-07-15 东南大学 Method for controlling fuzzy PID digital control DC-DC converter
CN107783423A (en) * 2017-10-25 2018-03-09 珠海格力电器股份有限公司 PID parameter self-tuning method and device based on machine learning
US20190187631A1 (en) * 2017-12-15 2019-06-20 Exxonmobil Research And Engineering Company Adaptive pid controller tuning via deep reinforcement learning
CN109254528A (en) * 2018-11-29 2019-01-22 曾喆昭 The fast wisdom PID control method of one kind three
CN111582493A (en) * 2020-04-15 2020-08-25 马鞍山师范高等专科学校 Management system for deep learning model training

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
徐萍等: "一种改进的PID参数整定方法" *

Similar Documents

Publication Publication Date Title
CN102427294B (en) Constant-time control circuit for switch type adjuster as well as switch type adjuster using control circuit
US8274266B2 (en) Switch mode power supply with dynamic topology
US8436592B2 (en) Voltage buck-boost switching regulator
US7948224B2 (en) Feedback controller having multiple feedback paths
EP1922802B1 (en) Peak charging current modulation for burst mode conversion
US9030184B2 (en) DC-DC converter and voltage conversion method thereof
US8232789B2 (en) System and method for providing linear buck boost transitions within a buck boost converter
CN103513685A (en) Current mode voltage regulator with auto-compensation
CN104038048A (en) Boost converter
US10020732B2 (en) Power converter having low power operating mode
CN103683869A (en) Switching power supply control circuit, switching power supply and control method of switching power supply
US6281668B1 (en) Switching regulator control circuit of PFM system and switching regulator thereof
US11245332B1 (en) Reference voltage control in a switch mode power supply
CN106899032A (en) Power optimization device and method for energy collection equipment
US10727743B2 (en) Systems and methods for enhancing dynamic response of power conversion systems
TWI506385B (en) Pulse width modulation power converter and control method
US11177737B2 (en) Switching control for power converters
CN113285595A (en) PID parameter setting system and control method of digital power supply based on machine learning
CN111509978A (en) DC-DC conversion device and power management system
CN115079763B (en) LDO circuit, control method, chip and electronic equipment
CN102684483A (en) Control circuit of switch-type adjuster and control method thereof
CN112051883B (en) Chip control method for realizing quick current response
CN202978705U (en) DC/DC output circuit
JP6593927B2 (en) Switching power supply
CN111953191A (en) DC-DC converter with steady state current limiting

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
CB02 Change of applicant information

Address after: 519000 2706, No. 3000, Huandao East Road, Hengqin new area, Zhuhai, Guangdong

Applicant after: Zhuhai Yiwei Semiconductor Co.,Ltd.

Address before: Room 105-514, No.6 Baohua Road, Hengqin New District, Zhuhai City, Guangdong Province

Applicant before: AMICRO SEMICONDUCTOR Co.,Ltd.

CB02 Change of applicant information
RJ01 Rejection of invention patent application after publication

Application publication date: 20210820

RJ01 Rejection of invention patent application after publication