Self-adaptive control method for intelligent bone drilling of mechanical arm

Abstract

The embodiment of the invention provides an intelligent bone drilling self-adaptive control method for a mechanical arm, which comprises the following steps: acquiring displacement deviation of the tail end of a drill bit of an orthopedic surgical robot, obtaining an error change rate through differential calculation, and combining a bone layer type as an input variable of a fuzzy controller and a feeding speed as an output variable; input variables are fuzzified through a membership function, and a fuzzy language variable set is obtained; according to the set and a preset rule base, constructing a multi-rule group structure comprising three types of scenes of free space, cancellous bone and cortical bone, and setting corresponding fuzzy rule subsets for different bone layers; designing a ''minimum-maximum'' rule activation and combination strategy for reasoning by adopting a Mamdani type fuzzy reasoning method according to fuzzification input and a multi-rule group structure, and generating a fuzzification output variable set; and defuzzification is performed through a gravity center method to obtain a continuous speed instruction, and dynamic regulation and control of the drill bit feeding speed are achieved. When approaching a penetration point, the device automatically decelerates and safely stops, so that the safety and accuracy of an operation are effectively guaranteed, and over-drilling injury is prevented.

Description

[Technical Field]

[0001] This invention relates to an adaptive control method for intelligent bone drilling of a robotic arm, which is applicable to the precise control and safe shutdown optimization of the robotic bone drilling process in orthopedic surgery. [Background Technology]

[0002] In orthopedic surgery, robot-assisted systems are gradually replacing traditional manual operations due to their high precision, automation, and stability, becoming an effective means to improve surgical safety and consistency. Among these operations, bone drilling is a crucial step, directly affecting the depth of bone penetration and the safety of adjacent tissues. Traditional surgery relies heavily on the surgeon's experience and tactile judgment; however, bone tissue is complex, varies in hardness, and lacks clear real-time feedback during surgery, easily leading to misjudgment and excessive penetration, damaging important nerves or blood vessels. Especially in special cases such as limb deformities, the non-standard anatomical structure significantly increases operational risks, demanding even more stringent precision in drill control. Therefore, enabling surgical robots to intelligently adjust the drill feed speed based on real-time feedback, especially ensuring safe stopping control near the penetration point, has become a pressing technical challenge.

[0003] To address the aforementioned issues, existing patents, such as Taiwan Patent Application No. TW202231250A, propose a method for controlling alveolar bone drilling and implant placement in dental implant surgery systems. This method receives a target depth input command and generates first and second motor drive signals to control the bone drill's advance and stop within a specific depth range, thereby achieving greater precision in drilling and implant placement. While this technology achieves depth control during implant advancement and triggers a stop mechanism upon reaching the target depth, improving operational accuracy and safety, its control logic relies on preset fixed depth values. It is primarily suitable for alveolar bone environments with homogeneous structures and known tissue properties, lacking a mechanism to perceive and respond to real-time changes in tissue characteristics (such as different bone layer densities) during drilling. This makes it difficult to adapt to the complex and variable drilling environment in orthopedic surgery. Especially when penetrating different bone layers, its advancement speed and control strategy fail to dynamically adjust according to errors, and no systematic adaptive control mechanism is introduced, easily leading to positional control inaccuracies or delayed stopping at tissue boundaries.

[0004] Therefore, although existing patents can achieve automatic stopping of the drilling process to a certain extent by preset depth, they lack a fuzzy control strategy that dynamically adjusts based on real-time error, error change rate and bone layer information. They have not formed a fuzzy control closed-loop mechanism that can adapt to various bone layer environments and achieve continuous speed adjustment and gradual stopping. This is the key problem that the existing technology has not solved. [Summary of the Invention]

[0005] In view of this, the present invention provides an adaptive control method for intelligent bone drilling of a robotic arm. By integrating fuzzy control technology with multi-rule group strategy, it achieves precise control of the bone drilling process of the orthopedic surgical robot and ensures that the drill bit stops safely when it approaches the penetration point.

[0006] This invention provides an adaptive control method for intelligent bone drilling in a robotic arm, comprising:

[0007] The end displacement deviation of the drill bit of the orthopedic surgical robot is obtained, the deviation rate is obtained by differential calculation, and the drill bit error, error change rate and bone layer type are used as input variables, and the drill bit feed speed is used as output variable. The input variables are mapped to fuzzy linguistic variables through membership functions to obtain the fuzzified set of input variables.

[0008] Construct a multi-rule group structure and set the corresponding fuzzy rule subsets;

[0009] Based on the fuzzified input variables and the multi-rule group structure, the Mamdani-type fuzzy inference method is adopted, and a "minimum-maximum" rule activation and merging strategy is designed for inference to generate a fuzzified set of output control variables.

[0010] Based on the fuzzy set of output control variables, the centroid method is used for defuzzification to convert the fuzzy quantities into precise speed commands.

[0011] According to the speed command, the feed speed of the orthopedic surgical robot drill is controlled, and it is safely stopped when it approaches the penetration point.

[0012] In the above method, the end displacement deviation of the orthopedic surgical robot drill is obtained, the deviation rate is calculated through differentiation, and the drill error, error change rate, and bone layer type are used as input variables, while the drill feed rate is used as the output variable. The input variables are mapped to fuzzy linguistic variables through membership functions, resulting in a fuzzified set of input variables, including...

[0013] Input variables are defined using Gaussian membership functions and trigonometric membership functions;

[0014] The Gaussian membership function is:

[0015]

[0016] The trigonometric membership function is:

[0017]

[0018] Where μ is the center of the Gaussian function, σ is the standard deviation, c is the center of the trigonometric function, w is the width, x is the input variable, exp is the exponential operation with the natural constant e as the base, and max is the function to take the maximum value;

[0019] For the drill bit error e, NVL represents "very far", NL represents "far", NM represents "medium distance", NS represents "relatively close", NSS represents "very close", and ZE represents "about to penetrate". The Gaussian membership function is used to represent the degree to which the drill bit error e belongs to the fuzzy sets NVL, NL, and NM; the triangular membership function is used to represent the degree to which the drill bit error e belongs to the fuzzy sets NS, NSS, and ZE.

[0020] For the error rate of change de, let Z represent "stationary state", S represent "slow speed", VS represent "very slow speed", M represent "medium speed", F represent "fast speed", and VF represent "very fast speed"; the Gaussian membership function is used to represent the degree to which the error rate of change de belongs to the fuzzy sets M, F, and VF; the triangular membership function is used to represent the degree to which the error rate of change de belongs to the fuzzy sets Z, S, and VS.

[0021] For the variable bone layer type, FREE represents "idling state", CANCELLOUS represents "cancellous bone", and CORTICAL represents "cortical bone". Triangular membership functions are used to represent the degree to which the bone layer type belongs to the fuzzy sets FREE, CANCELLOUS and CORTICAL.

[0022] By mapping the input variables to fuzzy linguistic variables through membership functions, we obtain the fuzzified input sets e′, de′, and bone type′, where e′ is the fuzzified drill bit error, de′ is the fuzzified drill bit error rate of change, and bone type′ is the fuzzified bone layer type.

[0023] The above method constructs a multi-rule group structure and sets corresponding fuzzy rule subsets, including...

[0024] Based on the fuzzified set of input variables and the preset fuzzy rule library, a multi-rule group structure is constructed, which includes three typical drilling scenarios: free space idling, cancellous bone region, and cortical bone region. For drilling characteristics under different bone layer types, corresponding fuzzy rule subsets are set.

[0025] Free space idling rule subset: When the drill bit is in free space and there is a large distance error between it and the target area, the system outputs a large feed rate to quickly advance the drill bit; as the error gradually decreases, the feed rate decreases accordingly; when the drill bit approaches the target bone surface, a lower feed rate is output.

[0026] Cancellous bone region rule subset: When the drill bit enters the cancellous bone region and the error is large, a faster or medium feed rate is output to balance drilling efficiency and bone adaptability; as the drill bit gradually approaches the target point, the feed rate gradually decreases; when approaching the penetration point, a stop command is output.

[0027] Regular subset of cortical bone region: Considering the dense and hard characteristics of cortical bone tissue, the system outputs a medium or slow feed rate when the error is large, so as to reduce the drilling load and enhance the system stability; as the drill bit gradually approaches the penetration point, the output speed is further reduced to a stop.

[0028] Based on the fuzzified input variables and the multi-rule group structure, a Mamdani-type fuzzy inference method is adopted. A "minimum-maximum" rule activation and merging strategy is designed for inference, generating a fuzzified set of output control variables, including...

[0029] For the input fuzzy variables e′, de′, and bone type′, the Mamdani-type fuzzy inference method is used to activate the matching rules in the rule base and calculate the antecedent membership degree through the "minimum" operation.

[0030]

[0031] The consequent membership function, activated by the antecedent, is obtained by using the "minimum operation" to activate the antecedent and the consequent membership function.

[0032]

[0033] Then, the fuzzy sets are merged using the "maximum" operation to obtain the fuzzified set of output control variables:

[0034]

[0035] Where n is the number of rules. Let the membership degree of the antecedent of the i-th rule be , μ represents the membership function of the original consequent fuzzy set of the i-th rule. consequent (v) is the activated output fuzzy membership function, μ V (v) represents the set of output control variables after fuzzification.

[0036] Based on the fuzzified set of output control variables, the centroid method is used for defuzzification, converting the fuzzy quantities into precise speed commands, including...

[0037] The formula for deblurring using the center-of-gravity method is:

[0038]

[0039] Among them, v k μ is the discrete value of the output velocity. V (v k ) represents the corresponding membership degree, and m represents the number of discrete points;

[0040] By calculating the defuzzification formula of the centroid method, the fuzzy set V is transformed into a precise speed command v, thereby achieving continuous control of the drilling speed.

[0041] According to the speed command, the feed speed of the orthopedic surgical robot drill is controlled, and it is safely stopped when approaching the penetration point, including...

[0042] The error between the drill bit position and the preset penetration point is monitored in real time. When the position error is less than the set threshold, the deceleration rule of the corresponding bone layer type is triggered through the multi-rule group structure, and the stop control rule in the fuzzy rule library is activated to control the robotic arm to gradually stop feeding, so as to ensure the end safety and control accuracy of the drilling process.

[0043] This invention integrates fuzzy control technology with a multi-rule group strategy, designing differentiated control rules for different bone layer characteristics (idling, cancellous bone, cortical bone). It leverages the advantages of Gaussian and trigonometric membership functions to achieve precise fuzzification of input and output variables, and utilizes Mamdani inference and the centroid method for defuzzification to generate continuous control commands. This method can dynamically adapt to changes in bone layer, significantly improving the accuracy of penetration point control while ensuring drilling efficiency. It avoids the delay risks associated with reliance on manual feel in traditional methods, ensuring surgical safety and stability, and providing an effective solution for the intelligent control of orthopedic surgical robots. [Attached Image Description]

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort or labor.

[0045] Figure 1 This is a fuzzy control principle diagram of an intelligent bone drilling adaptive control method for robotic arms provided in an embodiment of the present invention;

[0046] Figure 2 This is a graph showing the simulation results of an embodiment of the present invention.

[0047] Figure 3 These are simulation experiment animation diagrams of embodiments of the present invention. [Specific Implementation Examples]

[0048] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0049] It should be understood that the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0050] This invention provides an adaptive control method for intelligent bone drilling in robotic arms. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating an adaptive control method for intelligent bone drilling in a robotic arm, as provided in this invention example. Figure 1 As shown, the method includes the following steps:

[0051] Step 101: Obtain the end displacement deviation of the drill bit of the orthopedic surgical robot, obtain the deviation rate through differential calculation, and take the drill bit error, error change rate and bone layer type as input variables, and the drill bit feed speed as output variable. Map the input variables to fuzzy language variables through membership functions to obtain the fuzzified set of input variables.

[0052] Input variables are defined using Gaussian membership functions and trigonometric membership functions;

[0053] The Gaussian membership function is:

[0054]

[0055] The trigonometric membership function is:

[0056]

[0057] Where μ is the center of the Gaussian function, σ is the standard deviation, c is the center of the trigonometric function, w is the width, x is the input variable, exp is the exponential operation with the natural constant e as the base, and max is the function to take the maximum value;

[0058] For the drill bit error e, NVL represents "very far", NL represents "far", NM represents "medium distance", NS represents "relatively close", NSS represents "very close", and ZE represents "about to penetrate". The Gaussian membership function is used to represent the degree to which the drill bit error e belongs to the fuzzy sets NVL, NL, and NM; the triangular membership function is used to represent the degree to which the drill bit error e belongs to the fuzzy sets NS, NSS, and ZE.

[0059] For the error rate of change de, let Z represent "stationary state", S represent "slow speed", VS represent "very slow speed", M represent "medium speed", F represent "fast speed", and VF represent "very fast speed"; the Gaussian membership function is used to represent the degree to which the error rate of change de belongs to the fuzzy sets M, F, and VF; the triangular membership function is used to represent the degree to which the error rate of change de belongs to the fuzzy sets Z, S, and VS.

[0060] For the variable bone layer type, FREE represents "idling state", CANCELLOUS represents "cancellous bone", and CORTICAL represents "cortical bone". Triangular membership functions are used to represent the degree to which the bone layer type belongs to the fuzzy sets FREE, CANCELLOUS and CORTICAL.

[0061] By mapping the input variables to fuzzy linguistic variables through membership functions, we obtain the fuzzified input sets e′, de′, and bone type′, where e′ is the fuzzified drill bit error, de′ is the fuzzified drill bit error rate of change, and bone type′ is the fuzzified bone layer type.

[0062] Step 102: Construct a multi-rule group structure and set the corresponding fuzzy rule subsets;

[0063] Based on the fuzzified set of input variables and the preset fuzzy rule library, a multi-rule group structure is constructed, which includes three typical drilling scenarios: free space idling, cancellous bone region, and cortical bone region. For drilling characteristics under different bone layer types, corresponding fuzzy rule subsets are set.

[0064] Free space idling rule subset: When the drill bit is in free space and there is a large distance error between it and the target area, the system outputs a large feed rate to quickly advance the drill bit; as the error gradually decreases, the feed rate decreases accordingly; when the drill bit approaches the target bone surface, a lower feed rate is output.

[0065] Cancellous bone region rule subset: When the drill bit enters the cancellous bone region and the error is large, a faster or medium feed rate is output to balance drilling efficiency and bone adaptability; as the drill bit gradually approaches the target point, the feed rate gradually decreases; when approaching the penetration point, a stop command is output.

[0066] Regular subset of cortical bone region: Considering the dense and hard characteristics of cortical bone tissue, the system outputs a medium or slow feed rate when the error is large, so as to reduce the drilling load and enhance the system stability; as the drill bit gradually approaches the penetration point, the output speed is further reduced to a stop.

[0067] The specific fuzzy rules are shown in Table 1.

[0068] Table 1. Fuzzy rule table based on e′, de′ and bone type′

[0069]

[0070] Step 103: Based on the fuzzified input variables and the multi-rule group structure, the Mamdani-type fuzzy inference method is used to design a "minimum-maximum" rule activation and merging strategy for inference, generating a fuzzified set of output control variables.

[0071] For the input fuzzy variables e′, de′, and bone type′, the Mamdani-type fuzzy inference method is used to activate the matching rules in the rule base and calculate the antecedent membership degree through the "minimum" operation.

[0072]

[0073] The consequent membership function, activated by the antecedent, is obtained by using the "minimum operation" to activate the antecedent and the consequent membership function.

[0074]

[0075] Then, the fuzzy sets are merged using the "maximum" operation to obtain the fuzzified set of output control variables:

[0076]

[0077] Where n is the number of rules. Let the membership degree of the antecedent of the i-th rule be , μ represents the membership function of the original consequent fuzzy set of the i-th rule. consequent (v) is the activated output fuzzy membership function, μ V (v) represents the set of output control variables after fuzzification.

[0078] Step 104: Based on the fuzzy set of output control variables, the centroid method is used for defuzzification to convert the fuzzy quantities into precise speed commands.

[0079] The formula for deblurring using the center-of-gravity method is:

[0080]

[0081] Among them, v k μ is the discrete value of the output velocity. V (v k ) represents the corresponding membership degree, and m represents the number of discrete points;

[0082] By calculating the defuzzification formula of the centroid method, the fuzzy set V is transformed into a precise speed command v, thereby achieving continuous control of the drilling speed.

[0083] Step 105: According to the speed command, control the feed speed of the orthopedic surgical robot drill bit and stop it safely when it approaches the penetration point;

[0084] The error between the drill bit position and the preset penetration point is monitored in real time. When the position error is less than the set threshold, the deceleration rule of the corresponding bone layer type is triggered through the multi-rule group structure, and the stop control rule in the fuzzy rule library is activated to control the robotic arm to gradually stop feeding, so as to ensure the end safety and control accuracy of the drilling process.

[0085] Based on the method provided in the embodiments of the present invention, a simulation was performed to optimize the shutdown of the robot during orthopedic bone drilling operations.

[0086] The simulation experiment uses a negative coordinate system, with 0 as the initial drill bit position (negative for downwards), and the maximum drilling depth is set to -5cm. The environment includes three tissue types: air, cancellous bone, and cortical bone. Cortical bone is located between -2cm and -3cm (1cm thick), and cancellous bone is located between -3cm and -5cm (2cm thick). The stop threshold is set to -4.8cm. The initial drill bit position is 0, the initial velocity is 0, and the maximum velocity is limited to 5mm / s. The simulation runs in 0.1s time steps, recording the drill bit position, velocity, error (distance between the current position and the stop threshold), error rate of change, and control output in real time.

[0087] The fuzzy controller takes error, error rate of change, and bone layer type (determined by drill bit position) as inputs and outputs drill bit speed. The controller output is constrained by the maximum speed, and at each simulation step, it checks whether the stopping threshold has been reached. If the condition is met, the simulation stops and the final state is recorded.

[0088] Simulation results are as follows Figure 2 As shown, the drill bit feeds rapidly in the air at near maximum speed. After contacting the cancellous bone, the speed slows down, and further decreases upon entering the cortical bone. When approaching the stopping threshold, the speed gradually decreases until it stops. The speed in the air region is close to 5 mm / s, gradually decreasing to 2-3 mm / s in the cancellous bone region and to 1-1.5 mm / s in the cortical bone region. Within 5 mm of the threshold, the speed rapidly approaches 0. The control output is maintained at a high level in the air, gradually decreasing as the bone layer hardness increases and the error decreases, and then dropping sharply when approaching the threshold, dynamically matching the speed curve.

[0089] Visualization results as follows Figure 3 As shown, a safety deceleration mechanism is triggered when the drill bit approaches the stopping threshold, and the final position error with the threshold is less than 1 mm. The animation demonstrates that the drill bit automatically stops when it reaches the threshold. Simulation results show that the fuzzy control system can effectively control the movement of the drill bit in different bone layers, dynamically adjusting the feed rate according to the bone layer characteristics and error state to ensure a safe stop when approaching the penetration point. This verifies the feasibility and reliability of this control method in orthopedic surgery.

[0090] The technical solution of this invention has the following beneficial effects: This invention constructs a multi-rule group fuzzy control structure covering free space rotation, cancellous bone region, and cortical bone region, which can adapt to the drilling control requirements under various bone layer environments; by integrating drill displacement error, error change rate, and bone layer information, a multi-input variable fuzzification processing mechanism is designed, and Gaussian / triangular membership functions are used to improve the fuzzy mapping accuracy, enabling the control system to dynamically adjust the feed speed according to the drilling state; using a Mamdani-type fuzzy inference model, combined with a "minimum-maximum" rule activation and merging strategy to generate fuzzy output variables, and then defuzzifying them into continuous speed commands through the centroid method, smooth speed adjustment and stable control are achieved during the drilling process; in addition, by setting a safety threshold for approaching the penetration point and a bone layer-related rule triggering mechanism, the drill can automatically decelerate and gradually stop when it gradually approaches the endpoint, effectively reducing the penetration risk and ensuring surgical safety and accuracy.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0092] The contents not described in detail in this specification are common knowledge to those skilled in the art.

Claims

1. An adaptive control method for intelligent bone drilling in a robotic arm, characterized in that, The method includes: The end displacement deviation of the drill bit of the orthopedic surgical robot is obtained, the deviation rate is obtained by differential calculation, and the drill bit error, error change rate and bone layer type are used as input variables, and the drill bit feed speed is used as output variable. The input variables are mapped to fuzzy linguistic variables through membership functions to obtain the fuzzified set of input variables. Construct a multi-rule group structure and set the corresponding fuzzy rule subsets; Based on the fuzzified input variables and the multi-rule group structure, the Mamdani-type fuzzy inference method is adopted, and a "minimum-maximum" rule activation and merging strategy is designed for inference to generate a fuzzified set of output control variables. Based on the fuzzy set of output control variables, the centroid method is used for defuzzification to convert the fuzzy quantities into precise speed commands. According to the speed command, the feed speed of the orthopedic surgical robot drill is controlled, and it is safely stopped when it approaches the penetration point.

2. The adaptive control method for intelligent bone drilling of a robotic arm according to claim 1, characterized in that, The end-effector displacement deviation of the orthopedic surgical robot drill is obtained, and the deviation rate is calculated using differential methods. The drill bit error, error rate of change, and bone layer type are used as input variables, and the drill bit feed rate is used as the output variable. Membership functions are used to map the input variables to fuzzy linguistic variables, resulting in a fuzzified set of input variables. The specific process is as follows: Input variables are defined using Gaussian membership functions and trigonometric membership functions; The Gaussian membership function is: The trigonometric membership function is: Where μ is the center of the Gaussian function, σ is the standard deviation, c is the center of the trigonometric function, w is the width, x is the input variable, exp is the exponential operation with the natural constant e as the base, and max is the function to take the maximum value; For the drill bit error e, let NVL represent "very far", NL represent "far", NM represent "medium distance", NS represent "relatively close", NSS represent "very close", and ZE represent "about to penetrate"; Gaussian membership functions are used to represent the degree to which the drill bit error e belongs to the fuzzy sets NVL, NL, and NM; triangular membership functions are used to represent the degree to which the drill bit error e belongs to the fuzzy sets NS, NSS, and ZE. For the error rate of change de, let Z represent "stationary state", S represent "slow speed", VS represent "very slow speed", M represent "medium speed", F represent "fast speed", and VF represent "very fast speed"; Gaussian membership function is used to represent the degree to which the error rate of change de belongs to the fuzzy sets M, F, and VF; triangular membership function is used to represent the degree to which the error rate of change de belongs to the fuzzy sets Z, S, and VS. For the variable bone layer type, FREE represents "idling state", CANCELLOUS represents "cancellous bone" and CORTICAL represents "cortical bone". Triangular membership functions are used to represent the degree to which the bone layer type belongs to the fuzzy sets FREE, CANCELLOUS and CORTICAL. By mapping the input variables to fuzzy linguistic variables through membership functions, we obtain the fuzzified input sets e′, de′, and bone type′, where e′ is the fuzzified drill bit error, de′ is the fuzzified drill bit error rate of change, and bone type′ is the fuzzified bone layer type.

3. The adaptive control method for intelligent bone drilling of a robotic arm according to claim 1, characterized in that, The process of constructing a multi-rule group structure and setting the corresponding fuzzy rule subset is as follows: Based on the fuzzified set of input variables and the preset fuzzy rule library, a multi-rule group structure is constructed, which includes three typical drilling scenarios: free space idling, cancellous bone region, and cortical bone region. For drilling characteristics under different bone layer types, corresponding fuzzy rule subsets are set.

4. The adaptive control method for intelligent bone drilling of a robotic arm according to claim 1, characterized in that, Based on the fuzzified input variables and the multi-rule group structure, the Mamdani-type fuzzy inference method is adopted. Inference is performed using a "minimum-maximum" rule activation and merging strategy to generate a fuzzified set of output control variables. The specific process is as follows: For the input fuzzy variables e′, de′, and bone type′, the Mamdani-type fuzzy inference method is used to activate the matching rules in the rule base and calculate the antecedent membership degree through the "minimum" operation: The consequent membership function, activated by the antecedent, is obtained by using the "minimum operation" to activate the antecedent and the consequent membership function. Then, the fuzzy sets are merged using the "maximum" operation to obtain the fuzzified set of output control variables: Where n is the number of rules. Let the membership degree of the antecedent of the i-th rule be , μ represents the membership function of the original consequent fuzzy set of the i-th rule. consequent (v) is the activated output fuzzy membership function, μ V (v) represents the set of fuzzy output control variables.

5. The adaptive control method for intelligent bone drilling of a robotic arm according to claim 1, characterized in that, Based on the fuzzified set of output control variables, the centroid method is used for defuzzification to convert the fuzzy quantities into precise speed commands. The specific process is as follows: The formula for deblurring using the center-of-gravity method is: Among them, v k μ is the discrete value of the output velocity. V (v k ) represents the corresponding membership degree, and m represents the number of discrete points; By calculating the defuzzification formula of the centroid method, the fuzzy set V is transformed into a precise speed command v, thereby achieving continuous control of the drilling speed.

6. The adaptive control method for intelligent bone drilling of a robotic arm according to claim 1, characterized in that, Based on the speed command, the feed speed of the orthopedic surgical robot drill is controlled, and it is safely stopped when approaching the penetration point. The specific process is as follows: The error between the drill bit position and the preset penetration point is monitored in real time. When the position error is less than the set threshold, the deceleration rule of the corresponding bone layer type is triggered through the multi-rule group structure, and the stop control rule in the fuzzy rule library is activated to control the robotic arm to gradually stop feeding, so as to ensure the end safety and control accuracy of the drilling process.

Images