Electrosurgical radio frequency inverter based on model predictive control and frequency conversion control method thereof

Abstract

The invention discloses an electrosurgical radio frequency inverter based on model predictive control and a frequency conversion control method thereof, and the method comprises the steps: building a discrete output voltage prediction model based on a topological structure of the electrosurgical radio frequency inverter, and judging a switching frequency switching point and voltage peak output power through single sampling period iterative calculation. Constructing an output power prediction model; obtaining instantaneous prediction output power and theoretical reference output power obtained by real-time sampling and calculation when the output voltage is changed to a periodic peak value, and constructing a model prediction control evaluation function; according to the evaluation function, real-time voltage sampling and multi-frequency iteration prediction are combined with a set control target, the switching frequency of the radio frequency inverter is predicted and updated in real time, and constant-power frequency conversion control over the radio frequency inverter is achieved by continuously obtaining the control quantity at the next moment. The system can adapt to constant power output of a power supply under biological impedance change, can predict and control constant power output of an electrosurgical inverter in real time based on real-time updating of switching frequency, has good dynamic response performance, guarantees high efficiency and safety of clinical operations, and can be applied to various electrosurgical working platforms.

Description

Technical Field

[0001] This invention relates to the field of power electronic converter control technology, and in particular to an electrosurgical radio frequency inverter based on model predictive control (MPC) and its frequency conversion control method. Background Technology

[0002] An electrosurgical generator is a surgical device that uses high-frequency current to cut or coagulate biological tissue. Its core function is to provide stable and controllable radiofrequency energy. During surgery, the impedance of the load (i.e., biological tissue) will exhibit a wide range of dynamic changes due to variations in factors such as tissue type, humidity, and contact pressure.

[0003] Currently, many electrosurgical generators on the market employ traditional control strategies, such as proportional-integral (PI) control or fixed-frequency pulse-width modulation (PWM). These methods perform adequately when the load impedance is constant or changes slowly, but their dynamic response is sluggish when faced with rapidly changing bioimpedance, making it difficult to achieve precise constant power output. This can lead to uneven tissue heating, affecting surgical outcomes and even increasing the risk of thermal damage to healthy tissue. Therefore, there is an urgent need in the field for a control method that can rapidly respond to load changes and achieve high-precision constant power output. Summary of the Invention

[0004] The purpose of this invention is to provide an electrosurgical radiofrequency inverter based on model predictive control and its frequency conversion control method.

[0005] The technical solution to achieve the purpose of this invention is: an electrosurgical radiofrequency inverter based on model predictive control, comprising: a topology switching network module and a topology resonant network module;

[0006] The topology switch network module includes a first switch (Q1), a second switch (Q2), a third switch (Q3), and a fourth switch (Q4) for converting DC input voltage into high-frequency AC signals;

[0007] The topological resonant network module includes a resonant inductor (L... r ), resonant capacitor (C) r ) and transformer (T r ), used to achieve efficient transmission and conversion of radio frequency energy;

[0008] In this configuration, the source of the first switch (Q1) is connected to the drain of the second switch (Q2), the source of the third switch (Q3) is connected to the drain of the fourth switch (Q4), the drain of the first switch (Q1) and the drain of the third switch (Q3) are connected to the positive terminal of the DC input voltage, and the source of the second switch (Q2) and the source of the fourth switch (Q4) are connected to the negative terminal of the DC input voltage.

[0009] The resonant inductor (L) r One end of the transistor is connected to the node between the source of the first switching transistor (Q1) and the drain of the second switching transistor (Q2), and the other end is connected to the transformer (T). r One end of the primary winding is connected;

[0010] The resonant capacitor (C) r One end of the transformer (T) is connected to the transformer (T) r The other end of the primary winding is connected to the node between the source of the third switch (Q3) and the drain of the fourth switch (Q4).

[0011] The transformer (T) r The two ends of the secondary winding are connected in parallel with a load impedance (R). L ) and filter capacitor (C o ).

[0012] Furthermore, the topology switch network module is configured to receive the switching frequency control signal output in real time by the model predictive controller in order to dynamically adjust its switching frequency; the topology resonant network module is configured to adapt to dynamically changing bioimpedance under dynamic adjustment of the switching frequency in order to maintain constant output power.

[0013] A frequency conversion control method for the aforementioned electrosurgical radiofrequency inverter, characterized by comprising the following steps:

[0014] S1: Based on the topology of the electrosurgical radio frequency inverter, a discretized output voltage prediction model is established. Through iterative calculation in a single sampling period, the switching frequency switching point and peak voltage output power are determined, and an output power prediction model is constructed.

[0015] S2: Obtain the instantaneous predicted output power and theoretical reference output power at the peak of the cycle obtained from real-time sampling calculation of the output voltage change, and construct the model predictive control evaluation function;

[0016] S3: Based on the model predictive control evaluation function, the switching frequency of the RF inverter is updated in real time through real-time voltage sampling and multi-frequency iterative prediction to achieve constant power frequency conversion control.

[0017] Furthermore, in step S1, the specific process of constructing the output power prediction model includes:

[0018] S11: Based on the topology of the electrosurgical radio frequency inverter, a discretized output voltage prediction model is established;

[0019] Output voltage prediction value v o The formula for calculating (k+1) is:

[0020]

[0021] Among them, v o (k) represents the current output voltage value, a is the mode coefficient (a=1 when the inverter is in half-bridge mode, a=2 when the inverter is in full-bridge mode), and n is the transformer T. r Turns ratio; V in This is the DC input voltage; R L The load impedance at the output terminal; |Z total | represents the magnitude of the equivalent total impedance in the primary and secondary equivalent circuits; φ total f is the angle value corresponding to the equivalent total impedance. s T is the switching frequency of the switching transistor; k is the count value of the sampling period over a certain period of time. n For the sampling period, C o For filtering capacitors;

[0022] S12: Based on the topology of the electrosurgical radio frequency inverter, calculate the predicted output current corresponding to the branch where the load resistor is located, and derive the instantaneous prediction expression of the output power.

[0023] Output load resistor R L The expression for the predicted output current of the corresponding branch is:

[0024]

[0025] Among them, i RL (k+1) represents the predicted output current on the branch containing the output impedance, i o (k+1) is the predicted value of the secondary output current in the next cycle;

[0026] Instantaneous prediction of output power p o The formula for calculating (k+1) is:

[0027]

[0028]

[0029] Among them, v o (k+1) represents the predicted output voltage value at the next time step, v o (k) represents the current output voltage value, i o (k+1) is the predicted value of the secondary output current in the next cycle;

[0030] S13: By iterative calculation in a single sampling period, determine the effective value of the instantaneous predicted output power at the switching frequency switching point and the point where the output voltage reaches the peak output of the current period, and construct a complete output power prediction model;

[0031] In a single sampling period iteration calculation, the output voltage valley value of the current period is sampled to obtain v. o (0), and calculate the next sampling period T. n Output voltage value v o (T n ), and obtain v through iteration. o (2T n ), v o (3T n ), ...v o (kT n ), where k is the current sampling period number, k∈N * This process continues until the voltage value obtained through iteration first approaches zero, with an error range ≤0.707%V. o (0), at this point, the iteration point is determined to be the zero-crossing point of the output voltage waveform, and this moment is denoted as t. s ;

[0032] The switching frequency is switched to another switching frequency f at the zero-crossing point. si Continue according to the sampling period T n Iterative calculation of the voltage value v after switching o (t s +T n ), v o (t s +2T n ), ..., v o (t s +zT n ), where z is the iteration number. Due to the change in switching frequency, the time for the voltage waveform to rise from the zero-crossing point to the peak value changes. When the iteration reaches a certain voltage value and it is found that the voltage value in subsequent iterations begins to decrease, this point is the candidate frequency f. si The corresponding peak output voltage V peaki And i∈N * ;

[0033] Similar to the output voltage, the valley value of the output current in the current cycle is sampled to obtain i. RL (0), and the switched current value i is obtained through iteration. RL (t s +zT n The time when it reaches its peak value coincides with the time when the output voltage reaches its peak value, denoted as I. peakiIf and only if the instantaneous predicted output power p at the peak of the output voltage during this variation cycle is... o The effective value of (k+1) is expressed in the form of:

[0034] Among them, P oi The switching frequency switches to f when the output voltage crosses zero. si At that time, the predicted effective value of the output power is obtained.

[0035] Furthermore, |Z total The formula for calculating | is:

[0036]

[0037] Where, ω s For the corresponding switching angular frequency, and ω exists s =2πf s L r For resonant inductance, C r It is a resonant capacitor.

[0038] Furthermore, in step 2, the expression for the model predictive control evaluation function J(i) is:

[0039]

[0040] Where J(i) is the switching frequency that switches to f when the output voltage crosses zero. si At that time, the calculated value of the evaluation function, P o * This is the theoretical reference output power.

[0041] Furthermore, in step 2, the control objective set by the constant power frequency converter control is to minimize the model predictive control evaluation function J(i), that is, to obtain the instantaneous predicted output power p corresponding to the peak value of the output voltage in this variation cycle. o The effective value form P of (k+1) oi Approaching the current output power setpoint P o * (k).

[0042] The smaller the evaluation function, the smaller the deviation between the output power and the reference power in the next cycle. Therefore, model predictive control is transformed into solving for the optimal control quantity that minimizes the evaluation function. When the evaluation function approaches zero, the switching frequency f under optimal control can be obtained. s * .

[0043] Furthermore, in step 3, the specific process of updating the switching frequency of the RF inverter in real time is as follows:

[0044] S31: Preset effective adjustment range f of the switching frequency s ∈[f smin , f smax Frequency iteration step size ∆f s and voltage sampling period T n , where f smin To satisfy the minimum output power and the minimum switching frequency required for circuit resonance characteristics, f smax To avoid excessive switching power loss and electromagnetic interference, the maximum switching frequency is T. n It satisfies Shannon's sampling theorem and there exists T. n ≤1 / (2f smax );

[0045] S32: Based on the current switching frequency, in [f smin , f smax [Inside ∆f] s Generate a set of candidate switching frequencies;

[0046] S33: Based on the output power prediction model, the candidate switching frequencies f are calculated. si The predicted effective value of the output power P when the corresponding output voltage reaches its peak value oi (i=1,2,…,2m);

[0047] S34: Calculate the predicted output power values ​​P oi Substituting the model predictive control evaluation function J(k), the evaluation function values ​​corresponding to each candidate frequency are calculated, and the candidate switching frequency f corresponding to the minimum evaluation function value is selected. s * This serves as the switching frequency for the next control cycle, and the switching frequency of the topology switch network module is updated accordingly.

[0048] Furthermore, if multiple candidate frequencies correspond to evaluation function values ​​that are all minimum values, then the candidate frequency closest to the current reference frequency is selected as the switching frequency for the next control cycle; if the candidate frequency corresponding to the minimum evaluation function value exceeds [f smin , f smax If the range is defined, the frequency closest to the minimum evaluation function value at the boundary of that range is selected as the switching frequency for the next control cycle. If the candidate frequency is less than f... smin Then choose f smin If the candidate frequency is greater than f smax Then choose f smax .

[0049] Furthermore, in step 3, the output power deviation is controlled within ±5% of the output power setting value.

[0050] Compared with the prior art, the significant advantages of this invention are:

[0051] 1) High-precision constant power output: Through model predictive control, the optimal switching frequency can be calculated in advance, so that the output power closely tracks the set value and the power deviation is controlled within ±5%, adapting to a wide range of changes in bioimpedance.

[0052] 2) Fast dynamic response: The optimized iterative algorithm makes the time for a single frequency decision less than 100μs, which can quickly follow the dynamic changes in tissue impedance and ensure the real-time performance and stability of the surgical procedure.

[0053] 3) Smooth switching and high reliability: By setting the switching point of the switching frequency at the zero-crossing point of the output voltage, the inrush current caused by voltage change is effectively avoided, the switching stress is reduced, and the system reliability is improved.

[0054] 4) Good versatility: The method and topology described in this invention can be widely applied to various electrosurgical platforms that require precise energy control. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the topology of the electrosurgical radio frequency inverter described in this invention. Figure 2 This is a flowchart of the frequency conversion control method described in this invention. Detailed Implementation

[0056] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0057] like Figure 1 As shown, an electrosurgical radiofrequency inverter based on model predictive control includes a first switch Q1, a second switch Q2, a third switch Q3, a fourth switch Q4, and a resonant inductor L. r Resonant capacitor C r and transformer T r .

[0058] The source of the first switch Q1 is connected to the drain of the second switch Q2, the source of the third switch Q3 is connected to the drain of the fourth switch Q4, the drain of the first switch Q1 is connected to the drain of the third switch Q3, and the source of the second switch Q2 is connected to the source of the fourth switch Q4. The drain of the first switch Q1 is connected to the positive terminal of the DC input voltage, and the source of the second switch Q2 is connected to the negative terminal of the DC input voltage. The resonant inductor L... r One end is connected to the source of the first switch Q1 and the drain of the second switch Q2, and the resonant inductor L r The other end is connected to transformer Tr One end of the primary winding is connected, and the resonant capacitor C r One end is connected to transformer T r The other end of the primary winding is connected to the resonant capacitor C. r The other end is connected to the source of the third switch Q3 and the drain of the fourth switch Q4. Transformer T r The two ends of the secondary winding are respectively connected to the load impedance R L The two ends are connected; the load R L With filter capacitor C o The two ends are connected in parallel.

[0059] The electrosurgical radio frequency inverter topology comprises two main network modules: a topology switching network module and a topology resonant network module. The topology switching network module, consisting of four transistors (Q1, Q2, Q3, and Q4), serves as the core power conversion unit, converting the DC input voltage into a high-frequency electrical signal and simultaneously adjusting the switching frequency in real-time using model predictive control. The resonant inductor L... r and resonant capacitor C r and transformer T r A topological resonant network module is constructed, with the load R connected to the secondary side of the transformer. L (Simulated bioimpedance) and filter capacitor C o It can achieve efficient transmission and conversion of radio frequency energy, and with the help of model predictive control, it can adapt to the dynamic changes of biological impedance to ensure constant power output.

[0060] based on Figure 1 The electrosurgical radiofrequency inverter topology shown is illustrated, and a model predictive control-based frequency conversion control method for the electrosurgical radiofrequency inverter is implemented. The process is as follows: Figure 2 As shown, it includes the following steps:

[0061] Step 1, Build a prediction model:

[0062] Based on circuit theory, Figure 1 The topology shown is modeled. Discrete prediction models for the output voltage and output power from the primary-side switching network to the secondary-side load impedance are derived. This model establishes the mathematical relationship between the output at the next time step and the sampled values, circuit parameters, and control quantities at the current time step.

[0063] based on Figure 1 In the topological resonant network module of the electrosurgical radio frequency inverter, there exists:

[0064]

[0065]

[0066] Where, ω s For the corresponding switching angular frequency, and ω exists s =2πf s f s L is the switching frequency of the topology switch. r For resonant inductance, C r For resonant capacitor, Z Lr Z is the impedance corresponding to the resonant inductor. Cr This is the impedance corresponding to the resonant capacitor.

[0067] To facilitate the calculation of loop impedance, an equivalent loop is established from the secondary side to the primary side:

[0068]

[0069]

[0070] Among them, C o R is the output filter capacitor. L Z represents the load impedance at the output terminal, n is the turns ratio of the transformer, and Z... Ceq Z is the equivalent capacitive reactance corresponding to the filter capacitor. RLeq This is the equivalent impedance corresponding to the load impedance.

[0071] Therefore, the equivalent total output impedance, consisting of the equivalent filter capacitor and the equivalent load impedance, is calculated after the secondary side is converted to the primary side, and is denoted as Z*:

[0072]

[0073]

[0074] Calculate the equivalent total impedance in the equivalent circuit, denoted as Z. total :

[0075]

[0076] Substitute Z Lr Z Cr and Z * After considering all the expressions, we get:

[0077]

[0078] Z total Expressed in argument form, it can be denoted as:

[0079]

[0080] Among them, the amplitude |Z total | can be represented as:

[0081]

[0082] Angle value φ total It can be represented as:

[0083]

[0084] In the topology model of an electrosurgical radio frequency inverter, the DC input voltage V in After being controlled by the high-frequency switching of the switching transistor, it is converted into the AC voltage at ports ab in the topology:

[0085]

[0086] Wherein, the connection point between the source of the first switch Q1 and the drain of the second switch Q2 is node a, and the connection point between the source of the third switch Q3 and the drain of the fourth switch Q4 is node b. ab (t) represents the AC input voltage at port ab. The circuit has two operating modes, where a is the mode coefficient. When the drive signals of switches Q1 and Q2 are complementary and have a dead time, Q3 is always off and Q4 is always on, the inverter operates in half-bridge mode, corresponding to a=1. When the drive signals of switches Q1 and Q4, and the drive signals of Q2 and Q3 are complementary and have a dead time, the inverter operates in full-bridge mode, corresponding to a=2.

[0087] Based on the calculated relationship between voltage and current, the resonant network current i is obtained. Lr for:

[0088]

[0089] After the primary and secondary sides are transformed, the secondary side current i o for:

[0090]

[0091] For the output of the electrosurgical radiofrequency inverter topology, the following analysis and calculations were performed:

[0092]

[0093] Where Z is the total equivalent impedance of the secondary output terminal in the topology.

[0094] Therefore, after simplification, we get:

[0095]

[0096] Therefore, it can be obtained that at the output of the topology model, the filter capacitor C... o and load impedance R L The combined equivalent impedance Z is:

[0097]

[0098] Expressing the equivalent impedance Z at the output terminal in terms of angle, we have:

[0099]

[0100] The amplitude |Z| can be expressed as:

[0101]

[0102] The angle value φ can be expressed as:

[0103]

[0104] In the output circuit of the electrosurgical radio frequency inverter topology, due to the filter capacitor C o The existence of allows us to obtain the following relationship for the output current:

[0105]

[0106] Discretizing the differential part of the voltage here, we can obtain:

[0107]

[0108] Among them, T n The sampling period is the time step of digital control, which theoretically occupies a very short time, equivalent to an instantaneous state; v o (k+1) represents the output voltage value at time k+1, v o (k) represents the output voltage value at time k.

[0109] Furthermore, substituting the above equation, we get:

[0110]

[0111] Furthermore, after rearranging the terms, we get:

[0112]

[0113] Thus, the output voltage v at the current moment was initially derived. o (k) and the output voltage v at the next moment o The calculation relationship of (k+1).

[0114] Furthermore, substituting the previously derived i... o The expression exists:

[0115]

[0116] This allows us to obtain a complete output voltage prediction expression based on the topology parameters of the electrosurgical radio frequency inverter.

[0117] Because at time k+1, there exists:

[0118]

[0119] Among them, i o (k+1) is the predicted value of the secondary output current at time k+1.

[0120] Therefore, the load resistance R L The expression for the predicted output current of the corresponding branch is:

[0121]

[0122] Among them, i RL (k+1) represents the predicted output current corresponding to the load resistance.

[0123] Furthermore, the output power, as the power consumed by the load resistor, is expressed as follows:

[0124]

[0125] Furthermore, the output voltage valley value of the current cycle is sampled to obtain v. o (0), and calculate the next sampling period T. n Output voltage value v o (T n ), and obtain v through iteration. o (2T n ), v o (3T n ), ...v o (kT n ), where k is the current sampling period number, k∈N *This process continues until the voltage value obtained through iteration first approaches zero, with an error range ≤0.707%V. o (0), at this point, the iteration point is determined to be the zero-crossing point of the output voltage waveform, and this moment is denoted as t. s ;

[0126] The switching frequency is switched to another switching frequency f at the zero-crossing point. si Continue according to the sampling period T. n Iterative calculation of the voltage value v after switching o (t s +T n ), v o (t s +2T n ), ..., v o (t s +zT n ), where z is the iteration number. Due to the change in switching frequency, the time it takes for the voltage waveform to rise from the zero-crossing point to the peak value changes. When the iteration reaches a certain voltage value and it is found that the voltage value in subsequent iterations begins to decrease, this point is the candidate frequency f. si The corresponding peak output voltage V peaki And i∈N * .

[0127] Similar to the output voltage, the valley value of the output current in the current cycle is sampled to obtain i. RL (0), and the switched current value i is obtained through iteration. RL (t s +zT n The time when it reaches its peak value coincides with the time when the output voltage reaches its peak value, denoted as I. peaki The instantaneous predicted output power p at the peak of the output voltage during this variation cycle is predicted if and only if... o The effective value of (k+1) can be expressed as:

[0128]

[0129] Among them, P oi The switching frequency switches to f when the output voltage crosses zero. si At that time, the predicted effective value of the output power is obtained.

[0130] Step 2: Construct the evaluation function.

[0131] Define a model predictive control evaluation function J(i), quantify the deviation between the predicted power and the set power, and the control objective is to find the switching frequency that minimizes J(i).

[0132] To measure the accuracy of predicted output power in real time, an evaluation function needs to be established:

[0133]

[0134] Where J(i) is the switching frequency that switches to f when the output voltage crosses zero. si At that time, the calculated value of the evaluation function, P o * This is the theoretical reference output power.

[0135] The smaller the evaluation function, the smaller the deviation between the output power and the reference power in the next cycle. Therefore, model predictive control is transformed into solving for the optimal control quantity that minimizes the evaluation function. When the evaluation function approaches zero, the switching frequency f under optimal control can be obtained. s * .

[0136] Step 3: Real-time iteration and frequency update.

[0137] Based on the evaluation function, the switching frequency of the RF inverter is predicted and updated in real time by real-time voltage sampling and multi-frequency iterative prediction, combined with the set control target. By continuously obtaining the control quantity for the next moment, constant power frequency conversion control of the RF inverter is achieved.

[0138] S31: Parameter Initialization. Based on the topology characteristics of the electrosurgical radiofrequency inverter, the range of bioimpedance variation, and the daily operating power requirements, the effective adjustment range of the preset switching frequency is f. s ∈[f smin ,f smax ]. Where, f smin To satisfy the minimum output power and the minimum switching frequency required for circuit resonance characteristics, f smax The maximum switching frequency is set to avoid excessive switching losses and electromagnetic interference.

[0139] Furthermore, to balance iteration accuracy and real-time performance, the frequency iteration step size ∆f is set. s When the load impedance changes dynamically, through ∆f s Adjusting the switching frequency can keep the output power deviation within the clinically permissible error range (not exceeding ±5%).

[0140] Furthermore, the voltage sampling period is set to T. n It satisfies Shannon's sampling theorem, i.e., T n ≤1 / (2f smax This ensures that the sampled signal can accurately reproduce the characteristics of the output voltage waveform.

[0141] S32: Generate a candidate frequency set. Using the switching frequency f of the current cycle...s0 Based on the preset frequency range [f], smin , f smax Within ], according to the iteration step size ∆f s Generate all candidate switching frequencies, the set of which is denoted here as f. s ∈[f s0 -m·∆f s , f s0 -(m-1)·∆f s ,…, f s0 ,…, f s0 +(m-1)·∆f s , f s0 +m·∆f s ], where m is a positive integer, and it is necessary to guarantee f s All frequency values ​​are in [f smin ,f smax Within the specified range, avoid exceeding the circuit's safe operating range.

[0142] S33: Power Prediction. Based on the output power prediction model from step 1, the candidate switching frequencies f are calculated. si The predicted effective value of the output power P when the corresponding output voltage reaches its peak value oi (i=1,2,…,2m).

[0143] S35: Optimal Frequency Decision. The predicted output power P corresponding to each candidate frequency. oi Substituting (i=1,2,…,2m) sequentially into the model predictive control evaluation function J(k), the candidate frequencies f are calculated respectively. si The corresponding evaluation function values ​​J1, J2, ..., J 2m The core role of the evaluation function in this method is to quantify the deviation between the output power and the reference power. The smaller the J value, the stronger the effective value P of the predicted output power at that candidate frequency. o (k+1) The closer to the target power P o * (k) indicates a better control effect.

[0144] Therefore, for all the calculated model predictive control evaluation function values ​​J i Compare and select the minimum evaluation function value J. min = min (J1,J2,...,J 2m The corresponding candidate switching frequency is denoted as f. s * .

[0145] Furthermore, if there exist multiple candidate frequencies whose evaluation function values ​​are all equal to J... min(i.e., the minimum error values ​​are the same), then select the one closest to the current reference frequency f. s0 The candidate frequency is used as the final optimal switching frequency f. s * To avoid transient shocks to the circuit due to excessive frequency changes; if J min The corresponding candidate frequency exceeds [f smin ,f smax If the range is defined, then the point closest to J at the boundary of that range is selected. min The frequency as f s * This ensures that the circuit operates within a safe range.

[0146] The determined optimal switching frequency f s * The switching drive signal frequency for the next control cycle is output to the operating circuit of the electrosurgical radio frequency inverter via the switching transistor, enabling real-time updates of the switching frequency. Simultaneously, within the next control cycle, the above "sampling-iteration-calculation-filtering" process is repeated to continuously optimize the switching frequency, ensuring that the output power consistently tracks P. o * This enables stable constant power frequency conversion predictive control.

[0147] In summary, this scheme utilizes model predictive control based on the topology of an electrosurgical radiofrequency inverter to obtain predictive models for output voltage and power. It acquires inverter topology parameters, actual output voltage, actual output current, and output power setpoints to establish a model predictive control evaluation function. Based on this function, through real-time voltage sampling and multi-frequency iterative prediction, combined with the set control objectives, the switching frequency of the radiofrequency inverter is predicted and updated in real time. By continuously obtaining the control quantity for the next moment, constant power frequency conversion control of the radiofrequency inverter is achieved. This method can adapt to constant power output under varying bioimpedance conditions, exhibits good dynamic response performance, ensures the efficiency and safety of clinical surgery, and can be applied to various electrosurgical work platforms.

[0148] In summary, this invention can adapt to constant power output under changes in bioimpedance, retains the advantages of excellent transient performance of electrosurgical inverters, and can adapt to constant power output under changes in bioimpedance. By calculating the terminal current and terminal voltage, the interference of changes in bioimpedance on power calculation is solved. Based on the real-time update of the switching frequency, the constant power output of the electrosurgical inverter can be predicted and controlled in real time, ensuring the efficiency and safety of clinical surgery. It can be applied to a variety of electrosurgical work platforms.

[0149] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0150] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. An electrosurgical radiofrequency inverter based on model predictive control, characterized in that, include: Topology switch network module and topology resonant network module; The topology switch network module includes a first switch (Q1), a second switch (Q2), a third switch (Q3), and a fourth switch (Q4) for converting DC input voltage into high-frequency AC signals; The topological resonant network module includes a resonant inductor (L... r ), resonant capacitor (C) r ) and transformer (T r ), used to achieve efficient transmission and conversion of radio frequency energy; In this configuration, the source of the first switch (Q1) is connected to the drain of the second switch (Q2), the source of the third switch (Q3) is connected to the drain of the fourth switch (Q4), the drain of the first switch (Q1) and the drain of the third switch (Q3) are connected to the positive terminal of the DC input voltage, and the source of the second switch (Q2) and the source of the fourth switch (Q4) are connected to the negative terminal of the DC input voltage. The resonant inductor (L) r One end of the transistor is connected to the node between the source of the first switching transistor (Q1) and the drain of the second switching transistor (Q2), and the other end is connected to the transformer (T). r One end of the primary winding is connected; The resonant capacitor (C) r One end of the transformer (T) is connected to the transformer (T) r The other end of the primary winding is connected to the node between the source of the third switch (Q3) and the drain of the fourth switch (Q4). The transformer (T) r The two ends of the secondary winding are connected in parallel with a load impedance (R). L ) and filter capacitor (C o ).

2. The electrosurgical radiofrequency inverter according to claim 1, characterized in that, The topology switch network module is configured to receive the switching frequency control signal output in real time by the model prediction controller in order to dynamically adjust its switching frequency. The topological resonant network module is configured to adapt to dynamically changing bioimpedance under dynamic adjustment of the switching frequency in order to maintain a constant output power.

3. A frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 1 or 2, characterized in that, Includes the following steps: S1: Based on the topology of the electrosurgical radio frequency inverter, a discretized output voltage prediction model is established. Through iterative calculation in a single sampling period, the switching frequency switching point and peak voltage output power are determined, and an output power prediction model is constructed. S2: Obtain the instantaneous predicted output power and theoretical reference output power at the peak of the cycle obtained from real-time sampling calculation of the output voltage change, and construct the model predictive control evaluation function; S3: Based on the model predictive control evaluation function, the switching frequency of the RF inverter is updated in real time through real-time voltage sampling and multi-frequency iterative prediction to achieve constant power frequency conversion control.

4. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 3, characterized in that, In step S1, the specific process of constructing the output power prediction model includes: S11: Based on the topology of the electrosurgical radio frequency inverter, a discretized output voltage prediction model is established; Output voltage prediction value v o The formula for calculating (k+1) is: ； Among them, v o (k) represents the current output voltage value, a is the mode coefficient (a=1 when the inverter is in half-bridge mode, a=2 when the inverter is in full-bridge mode), and n is the transformer T. r Turns ratio; V in This is the DC input voltage; R L The load impedance at the output terminal; |Z total | represents the magnitude of the equivalent total impedance in the primary and secondary equivalent circuits; φ total f is the angle value corresponding to the equivalent total impedance. s T is the switching frequency of the switching transistor; k is the count value of the sampling period over a certain period of time. n For the sampling period, C o For filtering capacitors; S12: Based on the topology of the electrosurgical radio frequency inverter, calculate the predicted output current corresponding to the branch where the load resistor is located, and derive the instantaneous prediction expression of the output power. Output load resistor R L The expression for the predicted output current of the corresponding branch is: ； Among them, i RL (k+1) represents the predicted output current on the branch containing the output impedance, i o (k+1) is the predicted value of the secondary output current in the next cycle; Instantaneous prediction of output power p o The formula for calculating (k+1) is: ； ； Among them, v o (k+1) represents the predicted output voltage value at the next sampling time, v o (k) represents the current output voltage value, i o (k+1) is the predicted value of the secondary output current at the next sampling time; S13: By iterative calculation in a single sampling period, determine the effective value of the instantaneous predicted output power at the switching frequency switching point and the point where the output voltage reaches the peak output of the current period, and construct a complete output power prediction model; In a single sampling period iteration calculation, the output voltage valley value of the current period is sampled to obtain v. o (0), and calculate the next sampling period T. n Output voltage value v o (T n ), and obtain v through iteration. o (2T n ), v o (3T n ), ...v o (kT n ), where k is the current sampling period number, k∈N * This process continues until the voltage value obtained through iteration first approaches zero, with an error range ≤0.707%V. o (0), at this point, the iteration point is determined to be the zero-crossing point of the output voltage waveform, and this moment is denoted as t. s ; The switching frequency is switched to another switching frequency f at the zero-crossing point. si Continue according to the sampling period T n Iterative calculation of the voltage value v after switching o (t s +T n ), v o (t s +2T n ), ..., v o (t s +zT n ), where z is the iteration number. Due to the change in switching frequency, the time for the voltage waveform to rise from the zero-crossing point to the peak value changes. When the iteration reaches a certain voltage value and it is found that the voltage value in subsequent iterations begins to decrease, this point is the candidate frequency f. si The corresponding peak output voltage V peaki And i∈N * ; Similar to the output voltage, the valley value of the output current in the current cycle is sampled to obtain i. RL (0), and the switched current value i is obtained through iteration. RL (t s +zT n The time when it reaches its peak value coincides with the time when the output voltage reaches its peak value, denoted as I. peaki If and only if the instantaneous predicted output power p at the peak of the output voltage during this variation cycle is... o The effective value of (k+1) is expressed in the form of: Among them, P oi The switching frequency switches to f when the output voltage crosses zero. si At that time, the predicted effective value of the output power is obtained.

5. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 4, characterized in that, |Z total The formula for calculating | is: ； Where, ω s For the corresponding switching angular frequency, and ω exists s =2πf s L r For resonant inductance, C r It is a resonant capacitor.

6. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 3, characterized in that, In step 2, the expression for the model predictive control evaluation function J(i) is: ； Where J(i) is the switching frequency that switches to f when the output voltage crosses zero. si At that time, the calculated value of the evaluation function, P o * This is the theoretical reference output power.

7. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 3, characterized in that, In step 2, the control objective of the constant power frequency conversion control is to minimize the model predictive control evaluation function J(i), that is, to obtain the instantaneous predicted output power p at the peak of the output voltage in this variation cycle. o The effective value form P of (k+1) oi Approaching the current output power setpoint P o * (k) The smaller the evaluation function, the smaller the deviation between the output power and the reference power in the next cycle. Therefore, model predictive control is transformed into solving for the optimal control quantity that minimizes the evaluation function. When the evaluation function approaches zero, the switching frequency f under optimal control is obtained. s * .

8. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 3, characterized in that, In step 3, the specific process of updating the switching frequency of the RF inverter in real time is as follows: S31: Preset effective adjustment range f of the switching frequency s ∈[f smin , f smax Frequency iteration step size ∆f s and voltage sampling period T n , where f smin To satisfy the minimum output power and the minimum switching frequency required for circuit resonance characteristics, f smax To avoid excessive switching power loss and electromagnetic interference, the maximum switching frequency is T. n It satisfies Shannon's sampling theorem and there exists T. n ≤1 / (2f smax ); S32: Based on the current switching frequency, in [f smin , f smax [Inside ∆f] s Generate a set of candidate switching frequencies; S33: Based on the output power prediction model, the candidate switching frequencies f are calculated. si The predicted effective value of the output power P when the corresponding output voltage reaches its peak value oi (i=1, 2, …, 2m); S34: Calculate the predicted output power values ​​P oi Substituting the model predictive control evaluation function J(k), the evaluation function values ​​corresponding to each candidate frequency are calculated, and the candidate switching frequency f corresponding to the minimum evaluation function value is selected. s * This serves as the switching frequency for the next control cycle, and the switching frequency of the topology switch network module is updated accordingly.

9. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 8, characterized in that, If multiple candidate frequencies correspond to the minimum evaluation function value, the candidate frequency closest to the current reference frequency is selected as the switching frequency for the next control cycle; if the candidate frequency corresponding to the minimum evaluation function value exceeds [f smin , f smax If the range is defined, the frequency closest to the minimum evaluation function value at the boundary of that range is selected as the switching frequency for the next control cycle. If the candidate frequency is less than f... smin Then choose f smin If the candidate frequency is greater than f smax Then choose f smax .

10. The frequency conversion control method for the electrosurgical radiofrequency inverter according to claim 3, characterized in that, In step 3, the output power deviation is controlled within ±5% of the output power setting value.

Images